Start by taking stock of your old site

An old site usually has far more web addresses than you can see. Some stories dropped out of your menus years ago and still show up in Google or get visits from other websites. These are called orphan pages: nothing on your site links to them, but readers and search crawlers (the software Google uses to scan websites) can still reach them.

Old sites often have other surprises: addresses built on article ID numbers, dates in the path, pages created by plugins nobody remembers, content left over from a previous vendor, or a print-era archive imported years ago and never cleaned up. It is much easier to find them before the move than after.

Gather data from Google Search Console (Google's free tool that shows how your site performs in search), Google Analytics, your server logs, an SEO crawler (a tool that scans your site the way Google does) and the old CMS database. Together they show which pages actually bring traffic, which stories readers keep coming back to and which features your newsroom depends on. If your old vendor is slow to respond, ask for a full database and media export early.

Plan your URL changes

A URL is a page's web address. If a story has lived at one URL for years, collected links and ranked in search, giving it a new address means setting up a redirect, so that anyone who uses the old address lands on the new one.

The foundation is a redirect map: a list of every old URL next to its new one.

For example:

old URL: /2019/05/12/new-road-in-town

new URL: /local/new-road-in-town

After launch, the old URL should take readers straight to the matching story. If you are also deciding what your new URLs should look like, our guide to URL structure for SEO covers the basics: readable words, hyphens, folders instead of subdomains and canonical tags.

The usual approach is 1:1 mapping: each old URL points to one specific new URL. With a large archive this matters, because one wrong assignment can affect thousands of stories.

For a permanent move, Google recommends server-side redirects, including 301 or 308 (301 is the standard "moved permanently" code). Google also advises keeping redirects in place for at least a year.

Don't send hundreds of old stories to your homepage just to make "page not found" (404) errors go away. If a story has an equivalent on the new site, the redirect should go there. Google warns against mass-redirecting unrelated URLs to the homepage.

When a page was deliberately removed and has no equivalent on the new site, you can return a 410 ("Gone") status code.

Avoid redirect chains and loops

An old URL should point directly to the final URL rather than through another redirect. Redirect chains add extra requests and can become a problem on a large site.

A redirect loop is worse: address A redirects to B, and B redirects back to A. The browser keeps bouncing between the same two addresses, the story never opens and the reader sees an error.

Loops often appear when rules left over from the old system run alongside new ones, or when server rules conflict. This is common on older sites that have changed hands or vendors more than once.

Another frequent mistake is the soft 404: the page looks like an error ("not found"), but the server still reports success (a 200 code). To Google this is confusing: the page seems to exist, so it gets indexed even though it is useless to readers. In migrations it usually happens when an old URL with no equivalent is redirected to the homepage or a category page instead of returning a proper 404 or 410.

A site with a long archive can have thousands or tens of thousands of old URLs, so rules written one at a time don't scale. The redirect map should be prepared so the server can process it efficiently and the rules do not conflict with each other.

Keep in mind that 410 should not be the default answer of the whole server. It is a signal for individual pages you deliberately removed, not a bulk setting.

Move the full story data

Moving a story means more than moving its text. Keep the headline and title tag, the meta description (the summary shown in search results), publication and modification dates, the author, categories, tags, structured data (behind-the-scenes markup that helps Google understand a page), photos, galleries and internal links.

For newspapers it also means the things that make local news local: obituaries, legal notices, event listings, corrections and wire stories. For TV and radio stations it means embedded video, audio archives, podcast episodes and show pages.

Don't overlook the relationships between stories. If your old site shows related stories under an article, the latest items from a section or a most-read list, those features should keep working after the move.

What about stories published during the migration?

The newsroom keeps publishing during the migration. New stories, updates and photos still come in, and older stories may still be edited.

There are two common ways to handle this. With Delta-Sync, the whole database is copied at the start, and afterward only stories that were added or changed are synchronized.

With Dual-Publishing, selected stories are published in the old and new systems at the same time.

At switchover, the new CMS should hold current content, not only an archive copied a few days earlier. For a station or a daily paper, choose a switchover window away from your busiest news hours.

Test before you launch

The new version should first run in a test environment, so you can check everything without touching your live site.

Keep that version closed to the public and to Google. A simple option is Basic Auth: the server asks for a username and password. When you go live, remember to remove any test blocks (such as noindex tags or robots.txt rules), or Google will keep ignoring the new site.

On the test site, check URLs, redirects, canonical URLs, the sitemap, structured data, images, links and page layouts.

Google recommends testing the new version thoroughly before the migration begins and, when URLs change, preparing a mapping from old URLs to new ones.

Check the canonical URL of every page

The same story can sometimes be reached at several URLs. The canonical URL tells Google which version of a page you consider the preferred one, so canonical tags should be checked during the migration.

For the new version of a story, the right final URL should be set as canonical. When URLs change, Google recommends updating your internal links so they point to the new addresses. If you publish wire or syndicated stories, check which version should count as the original.

For more on this area, see our article on SEO fields.

Photos and images

Photos are often the most fragile part of a migration. Files get renamed, resized and re-exported, and the credits and captions that came with them are easy to lose along the way.

Credits can matter beyond courtesy. Under Section 1202 of the DMCA, intentionally removing or altering copyright management information, such as a photographer's credit, can create liability in certain circumstances. In Mango v. BuzzFeed (2020), the Second Circuit upheld a statutory damages award after a BuzzFeed journalist downloaded a photographer's picture from a newspaper's website, replaced the photographer's credit with a different name and used the photo without permission. In Murphy v. Millennium Radio Group (2011), the Third Circuit held that a photographer's credit printed beside a magazine photo can qualify as copyright management information, in a case where a radio station's website had used the image with the credit cut off. Section 1202 does not make every accidental loss of a credit a violation. The statute includes knowledge requirements tied to copyright infringement, and courts have interpreted those requirements differently. Still, it is worth preventing: store credits and captions as separate fields, and make sure resizing or re-exporting does not strip relevant IPTC metadata (the photographer and credit data stored inside the image file). This is general information, not legal advice.

Photo licenses should be reviewed before the migration. Wire pictures, stock images and freelancer work can carry limits on where, for how long and in what form they may be shown. Getty Images, for example, restricts its editorial images to editorial use and ties the terms of rights-managed editorial content to how, where and for how long they are used. Terms differ by agency and by contract, so go back to your own agreements with AP, Getty, other agencies and freelancers for time limits, archive rights and credit requirements. A migration is a good moment to record the source and license of every photo in the new system and to retire images you may no longer have the right to show.

If image URLs change during the migration, they should be redirected just like page URLs. Re-crawling images takes longer than pages, sometimes months, according to Google's office-hours guidance, so Google Images traffic may recover more slowly than page traffic. Keep the alt text (the short description of each image), and avoid placing key images only as CSS backgrounds, which Google Images does not index. If your old site passed creator, credit and license to Google through structured data or IPTC metadata, the new one should do the same.

Discover has its own picture rules: Google recommends large images at least 1200 px wide, the max-image-preview:large robots setting, and no site logo as the main image. Check that the new templates set this correctly, including the og:image of each story (the picture used when a story is shared or shown in feeds).

On TV and radio sites, the media library usually moves separately from the stories. Video poster images, show and host photos and podcast cover art are the files most likely to go missing.

Google News and Google Discover

For a news site, traffic from Google News and Google Discover (Google's personalized feed of stories in the Google app and on many phones) matters too.

The new system should keep the elements used for fresh content, including publication and modification dates, structured data, images and the News Sitemap, a special file that lists your newest stories for Google.

Google News has its own rules that matter in a migration. Its technical guidelines say article URLs must be unique and permanent, that you should not re-publish articles under a new URL, and that the URLs of your main news sections should not change often. If your URL pattern changes during a CMS move, Google asks publishers to send it their pattern transformation rules, and says it can help with these changes. Its crawler also reads HTML links best, so section pages should link to articles with plain HTML links, and the folders that hold your articles must not be blocked by robots.txt or meta tags.

More on the technical requirements is in our guide to Google News.

If Discover is an important traffic source, check it separately after launch. A migration alone doesn't guarantee that the traffic will hold, so watch the data in Search Console.

What it really costs, and who does the work

For a small newsroom, the price of a migration is not only the invoice. It is the staff hours spent on the move, the monthly fees for the old CMS, hosting, newsletter and e-edition tools, and the traffic you lose if something goes wrong.

Before you sign anything, ask: Is the archive migration included? Who builds and checks the redirect map? Is training included? And who picks up when something breaks on launch day?

For a local publisher, the practical questions are whether the newsroom can run the system without a developer and who will handle the migration and any problems after launch.

CMS4media and site migrations

CMS4media is a content management system (CMS) for local newspapers, TV and radio stations and online news websites. It combines the newsroom dashboard, page builder, SEO tools, advertising, e-editions, obituaries and multi-site management in one system. Editors can run the day-to-day work without a developer, with personal support from our team.

CMS4media runs local news sites across the United States, including Boerne Star, Taylor Press and Austin County Insider in Texas, Florida Weekly Destination in Florida, The Canton News in Mississippi and The Fallon Post in Nevada.

We have experience migrating sites with many years of archives. In these projects we handle the content migration, URLs, redirects and indexing, and we check the site after launch, so your newsroom keeps publishing instead of learning server rules.

In 2026, six news sites owned by Hartmann Media Consulting were moved to CMS4media. Their archives spanned years, so the migration also covered photos and metadata. After launch, the old URLs were audited, the 301 redirects were cleaned up, and indexing was checked in Google Search Console.

After launch, April was still a period of testing and of readers getting used to the changes. From June, results began to grow. Over the following months, traffic from Google Discover rose by nearly 41% and from Google News by 14.9%. Search traffic was maintained. The figures come from post-launch data reported in the case study and a conversation with the publisher.

If you use a paywall

If part of your content is for subscribers only, the access restriction has to survive the migration.

The new system should preserve the structured data that identifies paywalled content, including isAccessibleForFree and hasPart, so Google can understand which parts of the article require a subscription.

Check the performance of the new site

Changing the CMS can change how pages are generated, how the database is queried, how caching works and how ads are served.

A basic metric is TTFB (Time to First Byte): how quickly your server starts sending a page. Check real page load times too, and how the site behaves under heavier traffic, such as breaking news or election night. Stations should test the video and audio players as well.

For more on how page speed and your choice of CMS affect your SEO, read Speed Is the New Currency: Why Your CMS Choice Could Kill Your SEO.

If your site uses tools such as Prebid.js for header bidding, check the ad integrations during the migration, including how formats load and where ads appear.

For a larger site, a CMS migration can be a chance to tidy up ad operations and your work with advertisers.

Our publishers can use Ads4media, our platform that connects websites with advertisers. It works independently of the CMS, so it is available to publishers on other systems too.

In Ads4media, a publisher can present its advertising offer and receive orders for sponsored articles, banner campaigns and other forms of advertising. Publishers can manage these offers and orders in one place.

The CMS should fit how your newsroom works

The reason for a change may be performance, missing features, advertising problems, rising fees or too much dependence on developers.

In practice, a new CMS is judged by the daily work: how quickly editors can publish and update stories, how photos and galleries are handled, and how easily the site can grow.

When you run several sites

Each site can keep its own content, settings and history while sharing the same underlying system. This is useful for newspaper groups, chains of weeklies and groups of radio or TV stations.

CMS4media lets you manage multiple sites from one dashboard. The sites stay separate, while some tasks can be handled in one place.

What to check after launch

Once the new version is live, the most important things to watch are traffic, server errors and indexing (how many of your pages Google has added to its search results).

Compare your Google Search Console data with the period before the migration. Pay special attention to the number of indexed pages, 404 and 5xx errors (5xx means a server problem), and how your most important URLs behave.

Google recommends monitoring traffic on both the new and the old version of the site. In Search Console you can follow indexing, the sitemap and any errors that appear after the move.

Check the site on readers' phones and computers as well, including page load times.

Keep watching the site after the migration

Before work begins, analyze your current site, prepare the URL map, plan how to move the data and the stories published during the migration, and test the new version.

After launch, monitor traffic, indexing, server errors and performance.

The first days help you catch technical problems, but the fuller picture comes from the following weeks. Google notes that visibility can change temporarily during a migration, before the new URLs are processed. In the Hartmann Media Consulting case study, April was a period of testing and of readers getting used to the changes, and a clear rise in results was visible from June. It is a single example, not a rule, but it gives a rough timeframe to prepare for.

Over the following weeks, check whether the new URLs are indexed correctly, whether traffic is returning as expected and whether the CMS remains stable under real traffic.

Thinking about a move?

See CMS4media in the live demo, or talk to our team about your site, archive and migration timeline.

The ad appears. The revenue enters the report. Nobody has to call every advertiser individually or negotiate a separate campaign for each available space.

Eventually, the arrangement becomes infrastructure.

That is when changing it stops being a matter of replacing one line of code.

AdSense, DFP and AdX are related—but not identical

For many small websites, the Google advertising relationship begins with AdSense, which places ads and shares the resulting revenue with the publisher.

Larger or more complex publishers may use Google Ad Manager. One component of that system is the publisher ad server formerly known as DoubleClick for Publishers, or DFP. It decides which advertisement should fill an available space on a website.

AdX performs a different function. It is an advertising exchange where buyers compete for publishers’ inventory in real-time auctions.

Google owns both products.

The connection matters because the company also controls a major source of advertiser demand through Google Ads, previously called AdWords. A publisher choosing a different ad server or exchange risks losing access to part of that demand.

What looks like a collection of convenient tools can therefore become a tightly connected commercial system.

What the court found Google had done illegally

In April 2025, Judge Leonie Brinkema of the U.S. District Court for the Eastern District of Virginia found that Google had unlawfully acquired and maintained monopoly power in two markets for open-web display advertising:

• publisher ad servers;
• advertising exchanges.

The court found that Google had tied DFP and AdX together through contracts and technological integration. Publishers that wanted real-time access to demand from Google’s advertiser network were pushed toward using both Google products.

According to the court’s findings, Google also used policies and auction features that gave AdX advantages over competing exchanges.

By 2022, DFP served approximately 91% of open-web display impressions in the worldwide publisher ad-server market. AdX processed a majority of transactions among the exchanges that produced data for the litigation.

The court also found that Google had maintained a 20% AdX fee even as competing exchanges often charged closer to 10%. Publishers generally did not leave because access to Google’s advertising demand left them with limited practical alternatives.

That is the central antitrust problem.

The tools were not dominant only because publishers liked them. Google had designed the surrounding market in ways that made choosing a competitor economically difficult.

The Justice Department asked the court to separate the stack

After winning the liability phase, the Justice Department proposed structural remedies.

Google would have been required to sell AdX to an approved buyer. The company would also have had to open-source the final auction logic inside DFP. Under certain conditions, the remaining DFP business could have been sold later.

On September 2, 2026, the court rejected those structural remedies.

The judge did not reverse the earlier monopoly findings. She concluded that the proposed breakup was not a realistic or necessary way to restore competition when less disruptive behavioral remedies were available.

The evidence showed how deeply AdX depends on Google’s broader technical infrastructure. The government’s proposed schedule allowed as many as 540 days to migrate AdX to a buyer, followed by up to two additional years for customer migration. Identifying and approving a buyer would add more time.

Previous technology migrations described during the trial had taken even longer. Moving DoubleClick onto Google’s infrastructure took seven years. Replacing one dependency shared by AdX and DFP took five years and involved approximately 40 engineers.

The scale is equally difficult to separate. According to testimony from a Google engineering director, Google Ad Manager processes approximately 8.2 million ad requests and 60 million bid requests every second at peak.

A failed or degraded migration would not inconvenience publishers in the abstract. It could interrupt the system responsible for filling advertising space and producing revenue across thousands of websites.

The judge also found commercial uncertainty. The government had not identified a buyer clearly capable of acquiring AdX and reproducing its functionality at the required scale.

The breakup was rejected because the court considered it technically risky, commercially uncertain and broader than necessary—not because Google’s conduct had become lawful.

What Google must change instead

The tools will remain under Google’s ownership, but the ruling does not leave the existing arrangement untouched.

The court selected behavioral remedies intended to loosen the connection between Google’s products and lower the cost of choosing competitors.

According to the Justice Department’s description of the relief, Google will be required to:

• integrate AdX and DFP with Prebid, an open-source header-bidding system;
• allow AdX to submit real-time bids through competing publisher ad servers;
• let publishers access and export historical and configuration data from DFP;
• provide ongoing AdX bidding data, including winning and losing bids;
• prevent Google Ads from bidding preferentially into Google-owned tools;
• operate under a court-appointed monitor and technical committee for six years.

The Prebid integrations are especially important.

Instead of allowing DFP’s final auction to remain the invisible decision-maker, publishers may be able to run more of the competition through an auction whose logs and mechanics they can inspect.

Data portability addresses another source of dependence. A publisher that can export campaign settings, historical records and bidding information has a better chance of testing or adopting another ad server without rebuilding everything from the beginning.

These remedies do not create a competitive market overnight. They are designed to give competitors and publishers a better route into one.

Lock-in is measured by the cost of leaving

The case offers publishers a practical definition of vendor lock-in.

You are locked in when changing providers would require so much time, lost revenue, technical work and operational risk that staying becomes the only realistic choice.

The warning signs are not always visible when a system is first installed. They appear years later when the newsroom asks basic questions:

• Who controls our advertising accounts and historical data?
• Can we export our campaign settings in a usable format?
• Which demand sources disappear if we change ad servers?
• Who owns the relationships with our advertisers?
• How many tags, templates and consent settings would have to be rebuilt?
• Can we test another provider alongside the current system?
• How long could the site operate if the migration interrupted ad delivery?
• Does anyone in the organization still understand why the stack was configured this way?

If nobody can answer without calling the incumbent provider, the dependency is already part of the business model.

Publishers do not need to leave Google tomorrow

The lesson is not that every local newspaper should immediately remove Google advertising.

For many publishers, programmatic demand remains useful. A rushed migration can lose revenue, break reporting and create more operational risk than it removes.

The practical goal is optionality.

A publisher can maintain programmatic advertising while strengthening revenue streams it controls more directly. That includes locally sold campaigns, sponsorships, contextual placements, newsletter advertising, events and business services.

CMS4media, for example, supports multiple publisher-controlled advertising formats, including standard banners, contextual ads, website backgrounds, pop-ups, video and audio advertising. It can also accept external advertising code.

Those tools do not reproduce the scale of AdX, and they should not be described as a complete replacement for a global exchange.

They create a second path.

A campaign sold directly to a local hospital, dealership or contractor belongs to a relationship the publisher controls. The newsroom knows the advertiser, the price, the placement and the duration. That revenue does not disappear because an external auction changes its rules.

Build the exit before you need it

The court’s ruling exposes a difficult paradox.

Google’s products became so interconnected and important that dismantling them risked harming the same publishers the antitrust case was meant to protect.

That does not make the monopoly harmless. It demonstrates the cost of allowing dependence to accumulate for years before asking how the system could be separated.

Publishers cannot rewrite the advertising market by themselves. They can reduce their own exposure.

Document the current stack. Export the data that can be exported. Preserve direct advertiser relationships. Test additional demand sources. Make sure the website can support advertising that the publisher sells and controls.

The most useful question is not whether the newsroom could abandon Google next Monday.

It is whether the publisher is building enough alternatives to have a meaningful choice next year.

The cost may not appear on a credit card statement. It can show up when a county, school district or city borrows money to build a road, repair a water system or finance a public facility.

When lenders believe that nobody is watching how local government spends its money, they may price that additional risk into the loan.

That is the economic value of accountability reporting that pageviews do not capture.

What happens to borrowing costs when a newspaper closes?

A 2020 study published in the Journal of Financial Economics examined what happened to public finance after local newspaper closures.

Researchers Pengjie Gao, Chang Lee and Dermot Murphy found that municipal borrowing costs increased by approximately 5 to 11 basis points following a closure. They estimated an average additional cost of about $650,000 for each municipal bond issue.

The researchers also associated the loss of local reporting with higher government wages and deficits, as well as a greater likelihood of using more expensive refinancing and negotiated bond sales.

Their explanation was straightforward: local newspapers monitor public officials. When that monitoring disappears, lenders see a greater risk of inefficient spending or financial misconduct and charge accordingly.

The authors reported that the effect could not be explained simply by deteriorating local economic conditions causing both the newspaper closure and the higher borrowing costs.

Losing a newspaper did not only leave residents with fewer stories. It changed how financial markets assessed the community.

Where California’s $47.6 million estimate comes from

In June 2026, Rebuild Local News research director Matt Baker and study co-author Dermot Murphy expanded the original analysis.

They used a weighted average borrowing premium of 8.6 basis points, municipal bond issuance data and estimates of the population living in news deserts. Their model placed the additional nationwide borrowing cost associated with inadequate local coverage at approximately $1.1 billion annually.

For California, the estimate was $47.6 million a year.

That number needs a careful label.

It is a model of the cost associated with missing coverage. It is not a guaranteed saving that will appear automatically because California passed a law. The calculation depends on assumptions about which communities qualify as news deserts, how municipal borrowing is distributed and whether the effect identified in the earlier study applies across today’s bond market.

Restoring coverage would still have to improve scrutiny. Local governments would have to become more transparent or disciplined, and lenders would have to recognize the difference.

The estimate gives policymakers a way to describe the cost of losing local news. It does not eliminate the work required to rebuild it.

California has now put public money behind local reporting

On September 30, 2026, Governor Gavin Newsom signed AB 2222, the Community NEWS Act, into law.

Beginning with the 2027 tax year, qualifying California news organizations can receive refundable employment tax credits worth:

• $20,000 for each of their first five qualifying full-time journalists;
• $15,000 for each additional full-time journalist;
• an additional $15,000 for each new full-time journalism position;
• $7,500 for each qualifying part-time journalist.

Because the credits are refundable, eligible outlets can receive the balance even when the credit exceeds their tax liability. Print, digital and broadcast organizations can qualify, including nonprofits and sole proprietors that meet the law’s requirements.

The program also requires participating outlets to meet objective standards. These include disclosing ownership, maintaining a publicly available corrections policy and carrying media liability insurance. Political organizations and social-welfare groups cannot control eligible outlets.

We examined the structure—and the limitations—of the proposal earlier in The $20,000 Reporter. The important update is that the proposal is now law.

The credit applies to the cost of employing journalists. It does not evaluate whether an individual story produced measurable public savings, nor does it direct funding exclusively toward existing news deserts.

A tax credit can preserve reporting capacity. Accountability still has to happen.

A reporter creates value beyond the people who click

Publishers usually measure journalism through the audience they can see: subscriptions, pageviews, newsletter opens and advertising impressions.

Those measurements matter because they keep the business operating. But they capture only part of the value.

A resident who never reads the newspaper may still benefit when a reporter questions an inflated public contract. A family without a subscription may still pay less in taxes if regular scrutiny discourages waste. A local company may operate in a healthier market because public decisions are made more transparently.

This is an economic spillover: one organization pays the reporter, while the benefits extend to people who may never become customers.

That helps explain why the market has struggled to finance local accountability journalism. A newsroom cannot invoice every resident who benefits indirectly from its reporting.

Policies such as AB 2222 attempt to recognize some of that gap. They do not replace the need for subscriptions, advertising or other revenue. They acknowledge that the value of a reporter is not limited to the revenue attached to an article.

Make the impact visible

Local publishers should not wait for an academic study to explain their value.

When reporting leads to a corrected public bill, the cancellation of an unexplained expense, the release of a withheld document or a change in an official decision, document what happened.

That can mean:

• publishing a clearly labeled follow-up;
• linking the outcome to the original investigation;
• adding a short “What changed after publication” note;
• maintaining an internal log of corrections, savings and public responses;
• producing an annual report showing what the newsroom’s work accomplished.

This is not self-congratulation. It gives readers, advertisers, donors and policymakers evidence they can evaluate.

The original story and its result should also remain connected. The Related Articles module in CMS4media can display additional coverage based on shared categories or tags. Used with a consistent accountability or impact tag, it can help readers move from the first question to the eventual answer instead of leaving the outcome buried elsewhere in the archive.

One article may expose the problem. The sequence of reporting demonstrates the value.

The cost of a reporter is visible. So is the cost of losing one.

A salary appears clearly in a publisher’s budget. The value created by that salary is spread across better-informed voters, more cautious officials and public decisions that receive closer scrutiny.

That value becomes easiest to see after it disappears.

California’s new tax credit will test whether public support can help newsrooms retain and add reporters. The $47.6 million borrowing-cost estimate offers a reason to believe the return could extend far beyond the participating publications.

But publishers still have a role in proving the case locally.

Do not show the community only what the reporting cost.

Show what changed because somebody was there to ask the question.

You'll find the new options under Settings > Google Services, at the bottom of the tab. They're off by default, so you decide whether and when they appear. The section is called "Google News integration" and holds two independent toggles: "Enable the Add to favorite sources in Google button" and "Enable the Follow us on Google News button." The panel is bilingual, so the same settings show up in both Polish and English.

Where these buttons come from

Both buttons sit on top of features Google rolled out itself. Preferred Sources in Google Search launched in 2025, starting in the US and India, and on April 30, 2026 it went global across every language where Top Stories is available. On August 20, 2026 Google went a step further and gave publishers an embeddable button that readers click right inside an article, with no detour through Google's settings. That's the mechanism 4media now exposes as a ready-made switch. The second button, "Follow us on Google News," uses the long-standing follow feature you set up in Google Publisher Center.

What it does for your publication

When a reader marks your site as a preferred source, Google shows your content more often and higher up in Top Stories — with a "preferred" badge — and it can carry the same badge in AI Overviews and AI Mode for that reader. For a local newsroom, one detail matters: Google frames this as a tool for sites of every size, naming "niche local blogs" alongside global news desks in its own announcement. Google also reports that a reader who marks a site as a preferred source is about twice as likely to click through to it again. The button is built so you don't lose that reader: after the click, your site lands on their list and they return to the exact article they were reading. "Follow us on Google News" works alongside it, building a base of regular followers who get your stories in their Google News feed.

"Add to favorite sources on Google"

This is the publisher-facing side of Google's Preferred Sources. When a reader clicks the button under an article, your site joins their preferred-sources list and they go straight back to reading. On their next news-related searches, your articles surface more often in Top Stories, AI Overviews and AI Mode. The feature works globally, in every language Google Search supports.

Turning it on is a single step: flip the toggle, click Save. This button needs no ID — it goes live the moment you save.

"Follow us on Google News"

A signed-in reader follows your publication in Google News with one click, and your stories land in their personalized feed more often. There's one extra requirement here: the button needs your Google News publication ID. You'll find it in Google Publisher Center. If your title doesn't have a publication in Google News yet, register first — only then do you get the unique ID, which you paste into the "Google News publication ID" field and save.

Turning it on, step by step

1. Go to Settings > Google Services.
2. Under "Google News integration," flip the toggle for the button you want.
3. For "Follow us on Google News," paste the publication ID from Publisher Center.
4. Optionally, enable "Show buttons' description."
5. Click Save.

The buttons appear in the bottom-right corner of the article, below the author and publication date.

The "Show buttons' description" option

Switch it on and a short note shows up under the buttons, telling the reader why to click — for example, that it helps your content appear more often in Google search results and recommendations. It's a small thing that lowers the barrier: few people click something they don't understand.

Common questions

Do I need Publisher Center for both buttons?
No. "Add to favorite sources on Google" works without any ID. Publisher Center and a publication ID are needed only for "Follow us on Google News."

Does this guarantee more traffic?
No. It's a reader-driven signal and one input among many, not a position switch. But the more people add you as a source and follow you, the more often Google reaches for your content.

Can I enable just one button?
Yes. Each toggle works on its own.

How do I remove a button?
Flip the toggle off and save — it disappears from your articles.

The easiest place to start is "Add to favorite sources on Google": one click, one save, nothing else. Save the Google News follow button for the moment you have your Publisher Center publication ID on hand.

For a local news site, that can mean a reader learns what the city council decided without ever opening the story.

For years, search delivered the readers who drove pageviews, ad revenue and subscriptions.

As Google and AI assistants answer more questions directly, that bargain is changing. A news outlet can be used as a source, cited, even linked, and still never see the reader on its own site.

Is it worth chasing AI citations if they don't necessarily lead to a visit?

A citation isn't a visit

A Pew Research Center study found that when a Google results page included an AI summary, people clicked a traditional result in 8% of visits. Without a summary, they did so in 15%. Clicks on links inside the summary came in at about 1%.

A news outlet can appear in an answer and get very little traffic from it. Still, a citation can matter even without the click: someone asking about a simple fact has no reason to go to the source, but someone asking about causes, consequences or a comparison, or wanting to check the data, might.

A citation, a click, brand exposure and a conversion are four different things.

A citation tells you the outlet was used as a source. It doesn't tell you whether the reader visited, remembered the brand or became a subscriber, member or donor.

Each AI system works differently

Perplexity puts its sources right in the answer, so readers can see where a fact came from and click straight through to the story.

ChatGPT often builds an answer from many sources at once, and the reader may never click on any of them.

Google AI Overviews and AI Mode can also expand a user's query into related searches, a process often described as query fan-out. That means a publisher can be cited even when it does not rank prominently for the original query, because its reporting answers one of the related questions the system generates behind the scenes.

For newsrooms, that makes it worth monitoring the questions around a topic, not just the headline query itself.

Muck Rack's What Is AI Reading? tracked more than 25 million links cited by ChatGPT, Claude and Gemini. Journalism made up roughly a quarter of them, spread across more than 20,000 outlets, and it varied by question: prompts about industry trends cited journalism in 46% of responses, more than twice the rate of how-to and comparison questions. Axios ranked among ChatGPT's three most-cited domains in 13 of 17 industries, the only news outlet to reach the top three for any provider. The prompts focused on brands and industries, so they say little about local news.

Reddit, meanwhile, shows up near the top of many lists of most-cited sources, including 5W's synthesis of more than 680 million AI citations. So should a publisher open a Reddit account and post there solely to land in ChatGPT's answers? In the U.S., about a quarter of adults use Reddit, but only 9% of adults regularly get news there, compared with 38% on Facebook and 35% on YouTube. Local news is a different case: 52% of adults at least sometimes get local news from online forums like Facebook groups and Nextdoor.

In Google's AI Overviews, the answer appears directly in search results, so the reader may never open the source page.

In an Ahrefs analysis of 863,000 keywords, 38% of pages cited in AI Overviews ranked in Google's top 10 for the same query, down from 76% in the company's July 2025 study. Another 31% ranked between positions 11 and 100, and 31% outside the top 100.

A high ranking can still bring a click. The same page can also be used to answer a reader's question without a visit.

There is another reason AI visibility can vary: what a system already knows about a publisher or brand. A well-known outlet may have a much larger footprint across the web and across the data used to build AI systems. That can affect how easily the system recognizes the outlet, its authority and the topics it covers before it runs a new search.

Check questions, not just stories

Start with the questions your readers ask, rather than with the stories already getting the most traffic.

For a local news site, that might mean questions about local development projects, home prices, transit, or decisions by the city council or school board.

For a business publication: the state of a particular industry, company earnings, forecasts, or the effects of regulatory changes.

For a consumer site: choosing a product or service, how to file a claim, what things cost.

You don't need hundreds of queries. A few dozen covering the topics your newsroom cares about are enough to see where your site appears.

Use the wording readers actually use: a broad one (“What's the latest on the downtown redevelopment plan?”), one tied to cost or impact (“How will the new school budget affect my property taxes?”), a comparison (“How do the two candidates for mayor differ on housing?”), plus the follow-ups the chatbot suggests.

For each question, check whether your site appears, which story is cited, whether the reporting is used and credited correctly, and whether the answer gives the reader a reason to open the story.

The citation itself can be wrong. A Columbia Tow Center test from March 2025, now over a year old, found that eight AI search tools gave incorrect answers to more than 60% of queries asking them to identify the source of a news excerpt. The tools have changed since then, but the underlying risk, that a citation names the wrong source, has not been retested at that scale.

Check the same questions again later. The answer and the citations may have changed.

A site may be cited hundreds of times for basic facts and barely appear on the subjects its newsroom covers most closely.

None of that is within an AI system's control, but a publisher's own news CMS is: it decides how easily a newsroom can manage URLs, redirects, sitemaps and structured data across hundreds of stories.

Some stories are worth more than a click

A short brief based on a wire story gives readers what they need in a few sentences, and AI can summarize it just as easily.

Original reporting is different. What readers can't get from a short summary is the data, poll results, interviews, documents and reporting behind the story. AI can quote the finding or summarize the conclusion. The underlying data, documents and full interview are still on the publisher's site.

How publishers are responding

The Reuters Institute's report Journalism, Media, and Technology Trends and Predictions 2026 points to growing investment in original investigations and reporting, analysis and explanation, and video. The same publishers also report producing less content that can easily be reproduced elsewhere, including simple service journalism and some evergreen material.

Chartbeat data cited in the Reuters report show organic Google traffic to more than 2,500 sites falling 33% globally between November 2024 and November 2025, and 38% in the U.S. Publishers expect search traffic to fall a further 43% over the next three years. Not all of that decline can be attributed to AI Overviews.

Chartbeat data shared with Axios show that small publishers (1,000 to 10,000 daily pageviews, a range that includes many local news sites) lost 60% of their search referrals over two years, mid-sized ones (10,000 to 100,000) lost 47%, and large ones lost 22%. ChatGPT referrals grew more than 200% over the past year, but chatbots still account for less than 1% of publisher referral pageviews.

Beyond citations: what to measure

Microsoft Clarity data on more than 1,200 publisher domains found that traffic from AI platforms was still under 1% of all traffic, though it grew strongly over the study period. On some of the sites studied, that traffic led to meaningful conversions more often than traffic from traditional channels.

That figure doesn't describe the whole market. Ten thousand visits from search may produce no signups, and a hundred visits from AI may produce a few newsletter signups or subscriptions.

Those figures are useful context. Your own traffic, audience and conversion data will tell you whether AI referrals matter for your site.

Look at where those visits come from, which stories generate them and what visitors do next. Our guide shows how to set up an AI traffic report in Google Analytics 4.

The paywall complicates things

The most valuable stories are also the ones publishers most want to put behind a paywall.

A publisher wants AI systems to know it has reporting on a topic, but doesn't want to give away all its value for free.

The lede, basic facts or main thesis can remain public. Detailed data, documents, the full analysis or an exclusive interview can stay behind the paywall.

Free and paywalled content also needs to be marked correctly. Google has its own mechanisms for this, based on structured data.

Schema.org or paywall markup isn't a switch that guarantees AI will cite the material.

Likewise, robots.txt doesn't solve the whole problem. It controls which crawlers can access your site, but it doesn't give you control over how every AI system uses information it finds on the web.

The llms.txt file was intended to help AI understand a site. Large-scale research covering hundreds of thousands of domains suggests that major AI systems rarely rely on it. In other words, adding the file is not a substitute for making the actual site, content and source structure clear. We covered the standard in a separate article.

SEO or GEO?

GEO, or generative engine optimization (also called AEO), is a newer name for a problem publishers have dealt with for years.

SEO has always been partly about making content easy to find and understand, and making its source clear.

Now AI systems can use that content to build the answer itself.

For publishers, that still means clear headlines and subheads, specific answers, data, sources, bylines and a readable structure.

The article still needs to work for the reader first. That does not mean breaking stories into artificial “AI-friendly” chunks. Search and AI systems can work with naturally written content when the structure is clear. What matters is whether a reader, or a system extracting an answer, can quickly identify the key point.

Put the answer close to the top. Use descriptive subheads. Make the source of the reporting clear. Clear structure and visible bylines also make it easier for AI systems to identify where the reporting comes from.

Google itself says there are no special technical requirements for appearing in AI Overviews or AI Mode. For publishers, GEO is best understood as another way of thinking about search visibility, not a separate technical discipline that replaces SEO.

ChatGPT also has advertising: ads are running in the U.S., separately from the answer and its citations. Early results are mixed. Some advertisers report useful traffic and conversions, while others report low-quality traffic, weak conversion rates and difficulties with attribution and optimization. Discussions among advertisers on Reddit also include campaigns that generated clicks without meaningful conversions.

There is still no reliable industry benchmark for ChatGPT Ads. Performance appears to vary by audience, product and campaign, so clicks alone are not enough to judge the channel.

The goal is not more citations. It is more valuable citations.

A citation for a basic fact may never generate a visit. A citation to original data, an investigation or an exclusive interview can put a newsroom's work in front of a reader who has a reason to learn more.

The citations worth paying attention to are the ones tied to original data, documents, interviews and reporting done on the ground. The real question for a publisher is not how many AI citations it can get, but what it cites, for which questions, and whether those citations lead to meaningful audience behavior.

This feature allows you to add the same advertising code to multiple websites at once. Instead of adding the code separately to each website, you can apply it to all websites assigned to the currently selected business profile in Ads4Media.

Log in to Ads4Media and navigate to the [cms4media portals] module. Click [Tools] to open the dropdown menu, then select [Collective advertising codes].

You’ll be taken to a screen displaying all collective advertising codes added so far. If you haven’t added any codes yet, the list will be empty.

To add a new code, click the green [+ Add new collective advertising code] button in the top-left corner.

A new form will open with the following options: 

1. Title – An internal name used to help you identify the code in the list. For example, “Sellwild widget - header code”. 
2. Type – Select one of the available options from the dropdown list. Each option includes a description explaining where the code will be placed and when it will be loaded.
3. Ad Code - Here you paste your advertising code.

4. Media – Click [+ Choose] to open a pop-up window where you can select the websites to which the code should be added. Click the media you wish to choose - such entries will be marked with green frame. You can additionally use filtering options which can be helpful while looking for a particular medium. 
   
   Confirm selection by clicking the [Done, close window] button at the bottom. 

   
   
   Important: You can only select media assigned to your currently active business profile. If you are a member of multiple business profiles, switch to the appropriate one before adding codes to websites assigned to another company. 
    

5. Excluded URLs – Specify any URLs where the code should not be injected. Click [+ Add the URL] to create a new field for each URL you want to exclude. 
    

Once you’ve completed the form, click [Save] to add the collective advertising code. 

Adding collective ads.txt codes

This feature works similarly to collective advertising codes, but allows you to add ads.txt files to multiple websites at once.

Log in to Ads4Media and navigate to the [cms4media portals] module. Click [Tools] to open the dropdown menu, then select [Collective ads.txt codes].

You’ll see a list of all collective ads.txt codes added so far. If none have been added yet, the list will be empty.

To add a new code, click the + Add new collective ads.txt code button in the top-left corner.

 

A new form will open with the following options: 

1. Title – An internal name used to help you identify the purpose of the code in Ads4Media.
2. Content of the ads.txt file – Paste the contents of your ads.txt file into this field.
3. Media – Select the websites to which the ads.txt content should be added.
   
   Important: Only media assigned to your currently active business profile will be available for selection.

Once you’ve completed the form, click [Save] to add the collective ads.txt code.

No agreed rate. No list of the articles used. No explanation of why this month’s payment increased or decreased.

Only a number on the account.

That is roughly the position publishers face in Google’s emerging AI contribution pilot. It may establish an important precedent: content used to improve an AI answer can create value that deserves payment.

Before publishers treat that payment as dependable revenue, however, they need something more familiar.

They need an account they can audit.

What Google is testing

According to reporting by Digiday, Google has invited at least dozens of publishers and other website owners to participate in an early-stage “AI contribution” pilot.

Participants receive an additional earnings panel inside Google Search Console. It displays a monthly payment and some payment history.

Google reportedly makes payments when it determines that a publisher’s content has contributed meaningfully to an AI-generated response across products such as Gemini, AI Overviews and AI Mode.

The payment is based on Google’s assessment of value rather than a simple count of how many times the content was retrieved or displayed. Publishers can leave the pilot, but they currently receive little detail about how individual payments are calculated.

Google has publicly confirmed the broader experiment. In June, the company said it was testing new partnerships with websites whose content improves the freshness and factual accuracy of generative AI answers through grounding.

The public announcement did not explain the payout formula. Details about the Search Console panel and monthly earnings come from Digiday’s conversations with industry sources and participating publishers.

That distinction matters. The existence of the experiment is confirmed. Its economics remain largely hidden.

This is not another traffic program

For most of the internet’s commercial history, the exchange between Google and a publisher was based on traffic.

Google indexed a story. A reader clicked the link. The publisher gained a pageview and an opportunity to show advertising, collect an email address or sell a subscription.

AI answers disrupt that sequence.

A publisher’s reporting can help answer a user’s question without the user ever opening the original page. The reporting still contributed value, but the familiar commercial event—the visit—never occurred.

We recently examined this distinction in Is It Worth Chasing AI Citations?. A citation, a visit, brand exposure and a conversion are four different outcomes. A publisher can receive the first without receiving any of the other three.

Google’s pilot introduces a possible fifth outcome: payment without a visit.

For a small newsroom, that is worth testing. A council investigation or original data project may be useful across thousands of AI answers while producing relatively little referral traffic. If some of that value returns to the publisher, it can help finance the next piece of reporting.

The opportunity should not be dismissed merely because the first version is incomplete.

It should also not be confused with a functioning market.

A monthly number is not a useful ledger

A publisher cannot evaluate an AI licensing program from the payout alone.

Suppose a newspaper receives $800 in one month and $1,200 in the next. The increase appears encouraging, but what caused it?

Was one investigation used extensively? Did Google assign more value to election coverage, restaurant guides or public-safety reporting? Did overall AI usage increase? Did the payment rate change? Was content used in a different product or country?

Without story-level reporting, the publisher cannot tell.

The same problem becomes more serious when the payment falls. A decline from $1,200 to $600 could mean that the content was used less often. It could also mean that Google changed the formula, the product mix or its definition of a meaningful contribution.

Those explanations require different business responses. A single earnings number cannot distinguish among them.

At minimum, a useful publisher report should identify:

• which URLs or pieces of content contributed to AI responses;
• whether the content was retrieved, used for grounding, cited or displayed;
• the Google product in which the contribution occurred;
• the number or range of qualifying uses;
• the markets and languages involved;
• the rate or value assigned to those uses;
• any changes to the calculation since the previous period;
• adjustments, disputes or corrections affecting the payment.

The advertising industry would not accept a campaign report containing only the final invoice. AI licensing should eventually meet a comparable standard.

The industry is already designing a better receipt

The transparency gap is not unique to Google.

On October 2, the Standards for Publisher Usage Rights coalition, known as SPUR, released a content telemetry standard intended to report what happens when publisher content moves through an AI system.

The proposed reporting chain distinguishes among several events: retrieval, grounding, citation, presentation and engagement.

Those differences are commercially important.

A model may retrieve a story but never use it. It may use the reporting to construct an answer without citing the publisher. It may display a link that nobody clicks. Each event represents a different type of value and should not be collapsed into one unexplained number.

A shared standard would not determine what every article is worth. It could provide the evidence publishers and AI companies need before they negotiate that value.

Google has not committed to adopting the SPUR standard. Its existence nevertheless shows what publishers are beginning to ask for: not simply a check, but traceable information about the use behind it.

How a small publisher should test the offer

A small or mid-sized publisher may have little leverage to negotiate a separate licensing agreement. Participating in Google’s pilot could still provide early revenue and useful experience.

The safest approach is to treat it as an experiment with a written measurement plan.

Before joining, record a baseline:

• monthly search and Discover traffic;
• measurable referrals from AI platforms;
• the stories and topics most often cited by AI systems;
• revenue associated with those visits;
• the newsroom’s current publishing volume and coverage priorities.

During the pilot, save every monthly payment and dashboard change. Record any information Google provides about content usage, product updates or methodology. Compare the payments with changes in traffic and citations, while remembering that correlation does not explain the calculation.

Ask direct questions:

• Can the payment be connected to individual URLs?
• Does citation affect value, or is unseen grounding enough?
• How are corrections and updated stories handled?
• Can a publisher audit or dispute the calculation?
• Does participation change the right to negotiate other licensing agreements?
• What notice will Google provide before changing the formula?

Most importantly, do not use an unexplained payment to redesign coverage.

If payments rise during an election month, that does not prove the newsroom should produce more political content. If they fall after investigative reporting expands, that does not prove the investigation had less value.

Without story-level evidence, the number is income—not editorial intelligence.

Experimental income should fund experiments

Some publishers quoted by Digiday described Google’s approach as collaborative and welcomed the precedent of direct payment. Other industry sources characterized the early offers as too small to justify participation or worried that accepting them could weaken publishers’ leverage in future negotiations.

Both reactions can be reasonable.

The decision depends on the amount, the rights requested, the information provided and the alternatives available to that publisher.

For a small newspaper, even a modest payment can be useful. It might finance a public-records request, a freelance data project or a short reporting experiment.

It should not justify a permanent hire until the publisher understands how dependable the revenue is.

Before treating the income as recurring, look for several months of history, clear contractual terms, a predictable payment process and enough reporting to explain major changes. Build the budget around revenue the newsroom can reasonably forecast, not the most encouraging number displayed during a pilot.

A new customer is welcome.

A customer that decides the price, calculates the bill and provides no itemized account is still a customer the publisher must evaluate carefully.

Google’s pilot could become the beginning of a new market for journalism used by AI. But a market requires more than payments.

It requires a price, a record of what was purchased and enough information for both sides to decide whether the exchange is fair.

It can be a customer using artificial intelligence to reduce repetitive production work.

It can also be a supplier whose reporting adds valuable knowledge to an AI product.

Buying the first service should not require the publisher to surrender its bargaining position in the second market.

That distinction is missing from an unusual argument Washington has presented in one of the most consequential copyright cases of the AI era.

What the Justice Department told the court

On September 1, 2026, the U.S. Department of Justice filed a statement of interest in the consolidated copyright litigation involving OpenAI, Microsoft, The New York Times, book authors and other publishers.

The plaintiffs allege that copyrighted books and news articles were used without permission to train OpenAI’s models. OpenAI and Microsoft argue that the training is protected by fair use.

Under Section 107 of U.S. copyright law, fair use is assessed through four factors, including the purpose of the use, the nature and amount of the copyrighted material and the effect on the potential market for the original work.

The Justice Department’s filing supports the technology companies’ interpretation. It argues that training language models is highly transformative because the resulting systems perform functions different from those of the individual works used during training.

The government also frames American AI development as an economic and national-security priority. It warns that an unfavorable ruling could raise development costs, weaken smaller AI companies and give foreign competitors an advantage.

This is the administration’s position, not a court ruling.

Both sides have asked U.S. District Judge Sidney Stein to rule in their favor on key issues without a trial. As of this writing, the court has not decided whether OpenAI’s training practices qualify as fair use.

Washington’s argument about small publishers

The filing makes a second argument aimed specifically at the publishing market.

If AI companies must purchase licenses for training material, the government says, publishers with the largest archives could receive the greatest payments. Licensing costs could become a barrier that only the largest technology companies can afford, while large legacy media organizations collect most of the money.

At the same time, the filing argues that AI helps underfunded newsrooms compete.

One example is a local outlet using an image generator to create an illustration that might otherwise require a photographer or a licensed image. Other AI tools can assist with translation, transcription, research organization and routine editing tasks.

The concern about market concentration is legitimate.

Current AI licensing deals have largely favored national and international publishers with large archives, recognizable brands and the legal resources to negotiate. A licensing system designed only around volume could leave a weekly newspaper outside the room.

But that is an argument for building a licensing market that includes small publishers.

It is not an argument that their reporting has no compensable value.

The illustration is not the expensive part

An AI-generated illustration may lower the cost of presenting a story.

It does not produce the underlying journalism.

Someone still has to attend the school board meeting, read the budget, request the contract, call the superintendent and give every party an opportunity to respond. The newsroom pays for that work before there is anything for an AI model to summarize, analyze or learn from.

For a local newspaper, the expensive input is often not the image beside the article. It is the reporter’s time and the institutional knowledge required to understand what happened.

That reporting may also be unusually valuable to an AI system precisely because it cannot be found elsewhere.

A large national publisher offers scale. A local newspaper offers facts that may exist nowhere else: the result of a county vote, the name of a new school superintendent, the terms of a zoning decision or the history behind a local dispute.

Archive size is therefore an incomplete measure of licensing value.

Ten thousand versions of information available throughout the internet may contribute less unique knowledge than one original report from a town nobody else consistently covers.

Being an AI customer does not eliminate supplier rights

The government’s argument treats AI assistance and compensation for training material as if they were opposing interests.

They are not.

A farmer can buy a tractor and still charge for the crop. A photographer can use editing software and still license the finished image. A publisher can use an AI tool and still negotiate the terms under which its journalism contributes to someone else’s commercial product.

CMS4media’s AI Assistant, for example, can help editors prepare titles, introductions and SEO descriptions from reporting already entered into the system. It can reduce repetitive work while leaving factual verification and final editorial judgment with the newsroom.

That is a customer relationship. The publisher uses a tool to improve its own workflow.

Training or grounding an external AI product with that publisher’s reporting creates a different relationship. The AI company is using an input created and financed by the newsroom.

The fact that the publisher benefits from AI in one context does not settle what should happen in the other.

Affordable tools and fair compensation can coexist comfortably.

A court victory would not automatically help every weekly newspaper

The opposite oversimplification should also be avoided.

If the court rejects OpenAI’s fair use defense, that would not automatically generate a check for every local publisher whose work may have entered a training dataset.

Small outlets would still face practical barriers:

• determining whether their content was used;
• finding the party responsible for that use;
• negotiating without dedicated legal staff;
• understanding the rights covered by an agreement;
• verifying usage and payments;
• collecting sums too small to justify individual enforcement.

The experience of other media-support systems shows why distribution matters as much as the headline amount. As we observed in Australia Can Make Big Tech Pay. It Cannot Make Big Tech Bargain, a policy can require technology platforms to contribute without guaranteeing that independent local publishers receive meaningful individual agreements.

The same risk exists in AI licensing.

If each newspaper must negotiate separately with a global technology company, the largest publishers will continue to have the advantage. A workable system for local news may require collective negotiation, standardized contracts, transparent usage reporting or a clearinghouse that can distribute smaller payments efficiently.

The important question is not only whether licensing is required.

It is whether a local newsroom can participate without spending more on lawyers and administration than it receives for its work.

What fair participation could look like

A system designed for small publishers would need to reduce both the power imbalance and the transaction cost.

That could include:

• clear information about whether a publisher’s content was included in training or used to ground an answer;
• standard definitions of training, retrieval, grounding, citation and display;
• contracts written for publishers without in-house legal departments;
• collective or representative negotiation;
• payments that recognize original and locally exclusive reporting, not only archive volume;
• an accessible method for opting in, opting out or reserving specific rights;
• reporting that allows publishers to verify how compensation was calculated;
• a practical dispute process for missing or incorrect payments.

None of those mechanisms requires the court to decide that every use of copyrighted material is infringement.

Fair use is context-specific. Different methods of acquiring, storing, transforming and reproducing material may produce different legal outcomes.

But the government’s small-publisher argument does not resolve those questions. It simply assumes that access to inexpensive AI tools may be more valuable to a small newsroom than the ability to negotiate over its own reporting.

Publishers should not have to accept that trade without evidence.

Small publishers need a seat, not an exemption from payment

The Justice Department is right about one risk: a licensing market can be designed badly.

It could reward archive size, exclude new entrants and concentrate both payments and AI development among a handful of large companies.

The solution is not to pretend that small publishers benefit when their bargaining rights disappear.

A local newsroom can use AI to work more efficiently and still expect fair terms when its original reporting helps build another company’s product. It can welcome cheaper production tools without treating the reporter’s labor as a free input.

The real conversation should focus on participation.

How can a weekly newspaper know when its work has been used? How can it negotiate without mounting an expensive lawsuit? How should payments recognize unique local reporting instead of rewarding volume alone?

A cheaper illustration may help publish the story.

Someone still has to pay for the reporter who discovered there was a story to tell.

A paywall isn't just a way to lock an article anymore. It's an attempt to answer a harder question: what is a reader actually worth?

That question matters even more in a market where paying for news has never been the default. A 2025 Reuters Institute survey found that in most countries studied, only a small minority of people say they paid for any online news in the past year - the U.S. figure sits well below half. In other words: most readers, most of the time, aren't paying for news as a category.

And yet the market keeps building more paid-access models - from large digital subscriptions to small, specialized outlets built from day one around a mix of free and paid content. That's not really a contradiction. It shows publishers no longer treat "paying for news" as one single decision. They're looking for narrower, more specific reasons a reader might actually pay.

Readers are reluctant to pay for news in general. But they'll pay for something concrete: data, analysis, access to know-how they can't get anywhere else.

The problem doesn't start at the paywall

For a long time, publishers thought of the paywall mainly as a tool to grow subscriber counts. A reader hits a few articles, a prompt appears, some users buy access.

Piano's data across its publisher base points to a more interesting pattern. Between 2022 and 2024, traffic across the publishers it tracked grew only 5%.  Conversions - cases where a visitor actually bought a subscription - rose 14%. Subscription revenue rose 28%. Among sites that were losing traffic during that period, 54% still managed to grow their number of paid subscriptions.

So it wasn't about finding more users. It was about finding higher-value ones - readers more likely to convert.

The 2025 numbers make that gap even sharper. According to a follow-up Piano report, search traffic across the same publisher base fell 36%, but revenue fell by much less, because the readers who remained converted at a higher rate than before. Average new-subscription revenue per search visitor rose 31%.

Less traffic doesn't have to mean proportionally less money.

That's a meaningful shift, because for years, a publisher's success was measured mostly in users and pageviews. Increasingly, the more important question is how many of those users come back, how often they engage, and whether there's a reason for them to stay.

The most valuable reader isn't the one who clicks the most

The same Piano data shows a sharp split between two kinds of search visitors. The first are casual readers - people who land on a single article from a specific search and typically never return. They made up nearly half of all search traffic across the publisher base. But they accounted for only 0.3% of new subscription revenue, and their average revenue per reader was under 20 cents.

The second group are highly engaged readers - people who come back to a publisher regularly, read widely, and treat that outlet as a recurring source of information. This group is far smaller in raw numbers, but generates a disproportionate share of subscription revenue. Its share of search traffic changed far less in 2025 than the casual group's did - meaning it was largely insulated from the broader traffic decline.

In practice, a shift in search algorithms can look far worse in traffic charts than it does in revenue. Thousands of casual visits disappear, but a large share of them were never worth much in subscription terms to begin with.

For a publisher, the more important question stops being "how many users do we have?" and becomes "how many of them have a reason to come back?"

That shifts the focus from pure traffic acquisition toward relationship-building. A newsletter, a reader account, personalized recommendations, or a direct return visit can matter more than a single search click.

In Piano's 2025 data, direct traffic - readers typing in a URL directly or using a bookmark - overtook search traffic's share for the first time, and traffic from aggregators, including Google Discover, also grew.

Digiday: registration as the bridge to a subscription

A smaller, specialized publisher shows this shift clearly. Digiday is a New York-based trade publication covering media, marketing and advertising. It launched in 2008 with what Wikipedia's own entry on the outlet describes as "optional registration" for most of its history - readers could browse freely, register, or do neither.

In 2019, Digiday explained its own strategy shift directly to readers. Rather than putting up a hard paywall, it expanded its existing membership program, Digiday+. Non-members would be capped at a small number of free articles - typically around four - while members got unlimited access, original research, and event perks. The company noted it had built this into a profitable, roughly 80-person media business without venture capital backing.

The first currency doesn't have to be money. It can be consent to further contact.

This matters especially for trade and specialist media. If a reader works in marketing, advertising or media, their value doesn't have to come from how many articles they've read - it can come from simply being part of a specific professional audience that advertisers and business partners want to reach. That's part of why Digiday built its membership tier around research, events and daily briefings, not just article access.

Not every registered reader becomes a paying subscriber. But every registered reader stops being anonymous - and that's already a starting point for a longer relationship.

Google leaves publishers room to decide how much to show

Discussions about paywalls often get reduced to a choice between protecting content and staying visible on Google. In practice, the system is more flexible than that framing suggests.

Google lets publishers decide for themselves how much of a locked article a casual search visitor sees before hitting a paywall. A publisher might give every reader a handful of free articles per month before restricting further access - Google itself suggests a starting point of roughly 6 to 10 articles a month, recommending publishers start near the higher end and test how tightening that limit affects traffic and conversion over time. A publisher can also choose to show only an opening excerpt - a few sentences, sometimes called a lead-in - instead of cutting a reader off completely once they hit the limit. Google treats this as good practice, since even a short excerpt gives a reader some value and can build curiosity that helps conversion, rather than simply presenting a wall. Google doesn't mandate any particular approach here - it leaves the decision to the publisher, as long as paid content is properly labeled.

This also affects visibility in search results: the more of an article Google can actually read and index, the better it understands what the piece is about, and the easier it is to surface it to the right searcher.

Google also pays attention to the experience of readers who aren't subscribers yet. Its own documentation for publishers states that reader satisfaction starts dropping noticeably once a paywall blocks access more than roughly 10% of the time - and Google explicitly advises against approaching that threshold without testing first, since that's the point where a publisher risks losing readers who haven't yet had a chance to see the value of the content.

A paywall can be more effective when most readers never actually encounter it - and only run into it when they're genuinely interested in that specific piece of content.

For a publisher, that means distinguishing between someone who stumbled onto a single article and a reader who consumes content regularly. For the second group, the barrier can reasonably appear sooner, since the odds they'd be interested in a subscription are higher.

The paywall is also a technical problem

How a paywall is built has direct consequences for SEO.

Google requires publishers to clearly mark which part of a page is paid content. Structured data is the mechanism for that - specifically the property isAccessibleForFree. It tells the search engine that a piece of content is deliberately restricted, rather than simply hidden from readers while still being served to the crawler - which would otherwise look like deception.

In practice, this is straightforward to set up. A publisher wraps the paid portion of an article in its own CSS class - say, .paywall - and then references that same class in the article's structured data. From that point on, Google knows exactly where the restricted content starts and ends, without having to guess.

Problems start when that labeling stops matching reality. Content that's actually free to everyone shouldn't be marked as paid, and vice versa - and the same labeling has to work identically across every version of the page, including mobile.

There's a more fundamental rule underneath all of this: a crawler and a reader have to see the same page.

If the full article isn't meant to be available before login, the publisher should choose an implementation that simply doesn't send that content to the browser at all - rather than sending it and hiding it with CSS. That second approach is called cloaking: Googlebot gets a different version of the page than an ordinary visitor, and Google treats that as a policy violation regardless of the publisher's intent.

A well-built paywall doesn't hide that content exists from Google. It restricts an ordinary reader's access to it.

After implementation, it's worth validating the markup - for example, with a rich results testing tool - and fixing any critical errors. Correct structured data is exactly what lets Google tell a legitimate paywall apart from cloaking, so a mistake here can cost an article its visibility entirely.

At large publishers, there's also the question of crawl budget. Thousands of new articles, category pages, filters and duplicate URLs create an environment where server performance and response time start to affect how efficiently a crawler can move through the whole site.

It's worth treating a paywall rollout as an SEO change, not just a business decision. That means separately tracking search traffic, individual article visibility, click-through rate, and what share of visits end at the paywall before a reader gets to read anything. Overly aggressive restrictions can hurt both user experience and search rankings - so it's better to roll changes out gradually and watch the effect than to set a limit once and forget about it.

The New York Times shows subscriptions can become the main revenue line

The U.S. market already has publishers for whom digital subscriptions aren't an experiment but one of the core pillars of the business.

The New York Times ended the first quarter of 2025 with 11.66 million net digital-only subscribers, after adding roughly 250,000 net subscribers in that quarter alone. Digital-only average revenue per user rose 3.6% year over year, driven largely by subscribers moving off promotional pricing onto higher-priced tiers. Growth in both subscriber count and revenue per user pushed digital subscription revenue up 14.4% year over year in that same quarter - and the company continued adding subscribers through every subsequent quarter of the year.

Digital subscription isn't a side project for a large publisher anymore. It's become the mechanism replacing the revenue print advertising used to provide.

That doesn't mean every outlet can copy this model. The Times has brand recognition, a large existing audience and scale that most publishers simply don't have. Digiday, described above, sits at the opposite end of that same problem - a smaller, specialized outlet for which a hard paywall would have been a risk, not an obvious choice.

A specialized outlet sells more than just an article

A general-news reader might show up for something they'll forget within days. A specialist publication can build value differently.

A reader might pay for data, analysis, expert commentary, tools, newsletters, or organized access to knowledge. The article becomes part of the product rather than the whole product.

The New York Times shows this even within a single subscription. Alongside news, the Times built NYT Cooking into a standalone paid product - a recipe database with step-by-step guides and its own app, launched free in 2014 and turned into a $5-every-four-weeks subscription in 2017, drawing over 10 million monthly users even before that shift. Today it sells separately for about $6 a month or $50 a year, but its bigger role is inside the Times' "All Access" bundle: executives have said repeatedly that subscribers who add Cooking, Games or other non-news products to their plan engage more, churn less, and pay more over time than a news-only subscriber does.

The more specialized the offering, the less sense it makes to treat a single article as the only thing being sold.

That's also an opening for smaller publishers. They don't need to compete with large outlets on volume or traffic if they can build a smaller but genuinely engaged audience around one specific, well-defined need - the way Digiday does within its own niche.

Not every publisher needs a complicated system

For the largest media groups, building sophisticated models that predict a reader's likelihood to convert can make financial sense. For a small outlet, the cost of that kind of system can outweigh the potential payoff.

That's part of why Google's Reader Revenue Manager is a useful option for smaller publishers - a tool that lets an outlet launch paid subscriptions or voluntary contributions without building its own payment infrastructure from scratch. Google takes a 5% transaction fee.

That lowers the barrier to entry considerably. A small newsroom can first test whether readers are willing to pay at all before investing in anything more elaborate.

For a small outlet, simplicity can be an advantage rather than a limitation.

 Substack is the clearest example of that kind of simplicity, and it's become impossible to ignore on the U.S. market. The platform lets a writer publish a newsletter with a free and a paid tier, handles billing through Stripe, and takes a flat 10% cut of whatever a writer charges. No separate infrastructure, no negotiation.

By early 2025, Substack had crossed 5 million paid subscriptions across its writers, with the platform's own creators collectively earning roughly $450 million a year in gross revenue. Some of that is concentrated at the very top. Historian Heather Cox Richardson's newsletter, Letters from an American, has more than 2.5 million subscribers and, by outside estimates, over 200,000 paying - enough to outearn most newsroom editors on a single email a night.

But most writers on the platform earn far less, and the typical paid publication charges around $5 a month or $50 a year - a price low enough that a reader can subscribe to several without much friction.

What Substack shows at scale, a lot of small U.S. newsrooms have started doing individually: skip the custom subscription system, charge a simple flat monthly or annual rate, and let the writing carry the sale.

Medium takes the opposite bet: one paywall, thousands of writers

Substack lets each writer set their own price and their own wall. Medium does the reverse - it runs a single, platform-wide paywall, and individual writers opt into it rather than build their own.

Readers get a handful of free articles a month; unlimited access costs $5 a month or $50 a year, with no ads anywhere on the site. A share of every reader's subscription fee is then split among the writers whose paywalled stories that reader actually engaged with - based on how long they read, not on traffic alone. A writer never sets a price or manages billing. They simply mark a story as members-only and let Medium's pooled subscription revenue do the rest.

That's a genuinely different answer to the same problem this whole piece has been circling: who decides what a reader pays for, and how directly is that tied to one specific piece of writing. Substack ties it directly - one writer, one price, one relationship. Medium pools it - thousands of writers sharing one reader base, with individual earnings shaped by engagement rather than a subscription of their own.

Neither model is "the" answer. They're two different bets on the same basic question: does a reader pay a writer, or pay into a pool that writers draw from?

What about the reader who won't pay?

Not every user becomes a subscriber. That doesn't automatically mean they stop being valuable.

Some publishers are experimenting with rewarded ads - a model where a reader unlocks a piece of gated content by watching a short ad instead of paying directly. In this setup, the money doesn't come from the reader at all. The publisher is selling an advertiser something else: a moment of attention and a chance to engage that reader directly.

That's an interesting direction precisely because it doesn't try to turn every reader into a subscriber.

A gentler version of the same logic shows up on platforms built around direct, voluntary support for creators, like Patreon. You're Wrong About, a journalism-driven history podcast, funds its production this way, drawing tens of thousands of supporters who pay a small monthly amount directly to the hosts - proof that with a genuinely engaged audience, this kind of model can support a real ongoing production team.

The hardest case is local news

The toughest challenge belongs to small, regional newsrooms. They don't have the scale of large outlets, their local ad market is limited, and they're often competing against free content produced by publicly funded local government sources.

In that situation, a hard paywall can be too risky. If an outlet doesn't already have a large audience and a strong brand, simply locking articles doesn't automatically create a willingness to pay.

That's one reason subscriptions are just one piece of a larger puzzle at small outlets. Alongside them, sponsored content - commissioned by local businesses and institutions, often through ad platforms that connect publishers with advertisers such as Ads4Media - sits next to grants, donations, and event revenue.

It's worth noting that the choice between paid access and registration doesn't have to mean building a separate, expensive system from scratch. Our CMS 4media includes a built-in Paywall and Registration Wall module, where a publisher decides which content to restrict, whether access requires payment or just registration, and how much material a reader sees before hitting that wall - without building any of it in-house.

A local publisher can't just copy a major newspaper's model. It has to assemble revenue from several smaller sources at once.

The biggest shift might come from search itself

Until now, Google has served publishers in two ways at once. It delivered traffic, and it helped readers discover a brand in the first place.

The growth of AI-generated answers complicates that arrangement. A reader can get an answer to their question without ever visiting the source site, and the publisher can lose the moment where it might have shown an ad, invited a newsletter signup, or nudged someone toward registration.

That threat has a concrete face now. In September 2025, OpenAI launched ChatGPT Pulse, a feature that researches overnight and delivers a personalized morning briefing. Each day it produces a handful of visual cards, built from a reader's chat history, saved memory, and - if connected - their calendar and inbox. It's explicitly modeled on the habit publishers have spent years trying to build: open an app first thing in the morning and get caught up on what matters to you.

One journalist covering the launch put the question bluntly: can this replace your morning newsletter? The difference is that Pulse's "newsletter" has exactly one subscriber. It's written by a model instead of an editor. And it cuts the publisher out of the loop entirely - no click, no ad impression, no registration prompt, nothing for a newsroom to convert.

There's a second problem buried inside the first one: what a reader gets back from an AI tool isn't always reliable, and few examples make that clearer than shopping. The Business of Fashion analyzed how ChatGPT, Claude and Gemini recommend sustainable clothing brands and found that a brand's online reputation and how a shopper phrased the question mattered more to the chatbot's answer than the brand's actual environmental record. In practice, that means a chatbot can recommend a brand that's simply good at talking about sustainability - a pattern regulators call greenwashing - over one that's actually doing the harder, less marketable work.

That finding cuts both ways. If a model is mostly reading what's already been written about a brand rather than verifying what a brand actually does, then being written about - clearly, repeatedly, in places a model is likely to pull from - stops being a nice-to-have for that brand. It becomes close to a requirement. A brand with no coverage, no reviews, no press, is effectively invisible to a chatbot regardless of how sustainable its supply chain actually is. Media coverage isn't just a shopper's discovery layer anymore; it's the training signal for how AI describes a company at all.

That matters for publishers too, not just shoppers. Specialist and trade outlets have spent years building the kind of editorial authority that's supposed to cut through exactly this kind of noise - vetting claims, checking sourcing, calling out greenwashing when they see it. If an AI tool skips that layer and answers straight from a model's read of brand reputation, the publisher doesn't just lose a click. It loses the one thing a subscription was supposed to be worth paying for in the first place: a source a reader can actually trust.

At the same time, Piano's data suggests that readers arriving via AI tools are currently few in number but potentially high in value. In December 2025, they accounted for just 0.23% of all conversions across the publisher base Piano tracks, but their average revenue per reader was $15.50 - compared to $3.36 for an ordinary search visitor.

The scale here is still small, so it's too early to call this a breakthrough. But the direction is worth watching, especially as tools like Pulse specifically target the daily-briefing habit that newsletters were built to own.

How many readers land on a page is becoming less important than the intent they arrive with.

That also changes what a paywall is for. A reader who lands on a specific, deliberate question can be worth more than someone who happened to open five articles off a homepage.

In that context, the newsletter-and-registration approach described above through Digiday's example gains extra weight. A publisher that knows a reader by name and email doesn't have to rely entirely on where that traffic came from - a classic search result, an AI-generated answer, or a briefing tool that never sends a reader to the source at all. It can simply reach that reader directly.

The paywall is no longer a wall

Publishing is shifting away from a simple free-or-paid choice toward multiple ways of monetizing the same attention. One reader might be monetized through ads, another through registration, a third through a subscription, a fourth through voluntary support, and a fifth through access to a specialized product - a paywall is just one of a broader set of ways to monetize a news site, alongside ads, affiliate networks, and sponsored content.

That also means there's no single correct paywall. A different model makes sense for a large general-news outlet than for a business publication, a specialized trade outlet like Digiday, or a small local weekly.

The shared problem stays the same: finding the right balance between how much content to leave open and how much is actually worth restricting.

The best paywall doesn't need to cover most of a site. It just needs to reach the right readers with the right offer - the ones who actually have a reason to pay.

That might be the most important shift happening in how online publishers make money today. A publisher isn't just fighting for the largest possible audience anymore - it's fighting for a relationship with a reader who has a reason to come back, register, and eventually pay.

And in a world where earning attention through search alone keeps getting harder, that relationship might be the most valuable asset a publication has.

1. You need to add a new C ad space, with the option “Available on lists, in between the elements” checked, and select the chosen lists.

Ads4media → cms4media portals → Options→ Ad spots→ Add new C ad space

To add an advertising placement to your portal, log in to the Ads4media.online panel and go to the [cms4media portals] module. Open Options and select Ad spots.

Then click the Add new C ad space button

Fill in the advertising placement name. You can add a description and set the maximum number of ads available in this advertising placement.

For the ad to be displayed between entries on lists, be sure to check the Available on lists, in between the elements option.

After selecting this option, you will be able to choose on which lists the ad should be available, after how many entries the first ad should appear, and whether it should repeat.

The first position on the list – after which entry the ad should appear, e.g. if the number 3 is entered in this field, the ad will appear between the third and fourth item on the list.

Repeat every X positions – after how many entries the ad should repeat, e.g. if the number 2 is entered in this field, the second ad will appear after two more entries from the previous ad.

2. The second step is adding advertising banner C

Ads4media → Adverts → Add new ad - Advertising banner → Localization C

In the Ads4media.online panel, go to the Adverts module. Click the Add new ad button in the upper left corner and select Advertising banner from the list.

Enter the ad name and be sure to select C advertising banner element in the Localization field.

Add the creative and fill in all the required fields to create the ad. Click Add new ad at the bottom of the page.

3. Launching the advertising campaign

Ads4media → Ad campaigns → Add new ad campaign 

The final step is adding an advertising campaign. In the ADS4media.online panel, go to the Ad campaigns module and click the Add new ad campaign button in the upper left corner.

Fill in the required advertising campaign settings, and then add an ad insertion. Select Advertising banner C: Website template element.

Select the ad from the list - the one that was added in step 2.

Also select the Ad spot - the one that was added in step 1.

After saving the advertising campaign, the ad will be displayed on lists between entries.

Those are important corrections to the original proposal. They also reveal a persistent missing middle.

An outlet below A$150,000 has a grant route. A large publisher has negotiating staff, legal resources and scale. Between them sit established local organizations that are too large for the smallest-publisher grant but still too small to bargain with a global platform on equal terms.

If a platform pays the charge, the statutory distribution formula may reach some of those outlets. If the platform pursues direct agreements instead, the multiplier makes them attractive but still does not give them a seat at the table.

Australia closed Meta’s escape route

The new law is a response to a weakness exposed by Australia’s earlier News Media Bargaining Code.

That code helped produce more than 30 commercial agreements, reportedly worth roughly A$250 million a year. Then Meta stopped renewing its Australian news deals and discontinued Facebook News in Australia in 2024. Under the earlier structure, removing news could also remove the reason to bargain.

The News Bargaining Incentive closes that exit. The charge is tied to covered digital advertising revenue, not to whether a platform chooses to carry news. A platform cannot make the obligation disappear simply by taking news off its service.

It can, however, choose to pay rather than negotiate.

That choice determines what the mechanism becomes in practice. If platforms sign a broad range of meaningful agreements, the law operates as an incentive for commercial cooperation. If they pay the charge and keep publishers at arm’s length, it operates more like a sector levy with a journalism distribution program attached.

Both outcomes produce funding. Only one produces a working relationship.

The U.S. is testing a different route

There is no national American equivalent to Australia’s new charge. Federal efforts to create a collective bargaining framework for publishers have repeatedly stalled, while states have moved toward programs that subsidize newsroom employment or civic information more directly.

New York’s Empire State Newspaper and Broadcast Media Jobs Program makes up to $30 million a year available through refundable credits for retaining and creating newsroom jobs. Half of the annual allocation is reserved for businesses with 100 or fewer New York employees.

Illinois uses a smaller but more tightly defined program. Its $5 million annual Local Journalism Sustainability Tax Incentive Program offers a $15,000 refundable credit for each qualifying journalist retained and another $10,000 for a new position. Eligibility turns on concrete local-news indicators: original public-interest reporting, a journalist working at least 30 hours a week and living within 50 miles of the coverage area, ownership disclosure and, for digital-only outlets, at least one community story per week plus an Illinois audience threshold.

Demand itself offers a lesson. Illinois reports that its 2026 allocation for retained-journalist credits is already fully committed, while credits for net-new positions remain available. An early implementation review found that smaller organizations could qualify, but paperwork, limited administrative capacity and payment delays still created barriers. Funding a job that exists and inducing a publisher to add a job are not the same policy problem.

California is considering another version. AB 2222, which reached the state Senate in an amended form in August, would provide refundable credits of up to $20,000 for each of a publisher’s first five qualifying full-time journalists, with separate amounts for additional, part-time and newly created positions. It is still a bill, not money a publisher can put in the budget. We examined the architecture and the caveat in The $20,000 Reporter.

The contrast with Australia is useful:

The American programs define the newsroom, employee or civic outcome first and then direct support toward it. That can target local reporting more precisely. Australia starts with the platform’s contribution and then creates two possible routes for the money: negotiated deals or statutory distribution.

The U.S. could eventually combine those strengths. A serious national model would guarantee a contribution, publish transparent distribution data, protect a meaningful share for independent and regional publishers, and let smaller outlets bargain collectively rather than approach global companies one by one.

The number of agreements should never be the only success metric. Policymakers and publishers need to know the value distribution, ownership profile, geography and newsroom employment supported by those agreements.

A payment buys runway. It does not build a business.

There is a temptation to treat any new policy dollar as the missing business model. It is not.

A platform agreement, employment credit or grant can keep a reporter on payroll. It cannot make readers return, convert anonymous traffic into a known audience, sell a local advertising campaign or turn strong reporting into recurring revenue. Public policy can lower the cost of journalism. The publisher still has to build the other side of the equation.

That is why the operational lesson matters as much as the legislative one. We made the same point in The $129 Million Scaffolding: institutional support rewards publishers that can document what they are, execute a campaign and handle new funding without creating a workflow bottleneck.

The infrastructure should also help the newsroom outlast the support. A Paywall and Registration Wall can turn selected reporting into subscription revenue or first-party reader relationships while leaving other coverage open. Within CMS4media, that capability sits inside the same publishing environment as the content it is meant to sustain.

The point is not that every local outlet should lock every article. It is that a temporary inflow should finance permanent capacity: better reporting, a stronger owned audience, cleaner reader data and more dependable revenue.

What American publishers should watch next

Australia’s headline number will be easy to follow. The more revealing evidence will take longer.

Watch how many platforms choose direct deals instead of paying the charge. Watch how much of the deal value reaches independent and regional publishers—not merely how many names appear on a list. Watch whether the small-publisher grants and the statutory distribution arrive quickly enough to affect newsroom employment. And watch how much disclosure the system produces around confidential commercial agreements.

Most of all, watch the missing middle. If established local publishers still cannot enter meaningful negotiations, the 200% multiplier will have improved the math without fixing the market.

Australia can now make Big Tech contribute to journalism even when a platform does not carry news. That is a genuine policy advance, and American lawmakers should study it closely.

But payment is an input. Distribution is a policy choice. Sustainability is still a business model.

This is the most popular type of advertising. On portals running on CMS4media, banners can be added in JPG, PNG, GIF and HTML5 formats. The maximum dimensions depend on the place where the banner is displayed.

Where can advertising banners be displayed?

• Top Banner – in the main part of the website, above the header
• Main Banner A – below the information bar
• ASUB Banner – below the main banner
• Banner E – above the portal footer
• Banner D – in the article:
  • Below the title
  • Below the introduction (lead)
  • After the first paragraph of the content
  • After the second paragraph of the content
  • After the third paragraph of the content
  • After the entire article content
• Banners B and C placed anywhere on the page as widgets
• Banner C between entries on lists (e.g. on the article list, event list, etc.) - learn more.

Website background

Website background surrounds the main content of the site, filling its side margins and the upper part of the screen.

The ad is very clearly visible and immediately attracts the attention of portal users, but it does not cover the published articles. Thanks to this, the publisher can offer business partners a highly visible, large advertising space without compromising readers' browsing comfort.

Top Layer (Pop-up)

Top Layer is displayed directly on top of the page and temporarily covers the main content of the site. Thanks to this, it perfectly focuses the portal user's attention on itself. On the other hand, overusing this format may make using the site cumbersome, which is why it is worth setting limits regarding the ad display. In CMS 4media you can decide:

• when the ad will be displayed - e.g. immediately after entering the page, after a specified number of seconds, or after scrolling halfway down the page,
• after how many seconds the ad will disappear
• after how many days, hours or minutes the user will see the ad again.

Video ad before video content

This is an ad displayed before editorial video content, highly engaging and visible to users. To make sure it is not too irritating for viewers, it is a good idea to set the number of seconds after which it can be skipped.

In CMS 4media, the available video ad formats are:

• in MP4 format
• generated by VAST code

Audio ad before audio content

The publisher has two formats to choose from when adding an audio ad:

• MP3
• VAST code

Like a video ad, it is played before editorial content and can be skipped after a few seconds. 

Contextual advertising

Contextual advertising is displayed when a specific keyword appears in the article content. The thematic match makes the campaign more effective.

Ads generated by code added in the <HEAD> / <BODY> section

In addition to standard advertising formats, the publisher of a portal running on CMS 4media can add external HTML code that triggers ads directly in the page code. 

The code in the <HEAD> section will be added just before its closing tag and, depending on needs, may be active:

• throughout the entire portal
• only on the homepage
• on all pages except the homepage
• on all pages (including the homepage) except pages with articles marked as sponsored.

In the case of code in the <BODY> section, the publisher can decide whether the code will appear:

• immediately after opening the <BODY> section
• immediately before its closing tag.
• below the article content;
• below the related articles section.

The robots.txt file is a simple text document that serves as an official instruction for search engines (such as Google). 

Before the algorithms of these search engines begin scanning a website, they first check this file to find out which sections they may index and which they should skip. Thanks to this, the publisher gains control over what appears in public search results. 

Using the robots.txt file, you can block access to selected subpages, such as the admin panel, draft versions of articles, pages under construction, or private website files.

Robots.txt is based on two basic directives:

• User-agent: Specifies which search engines a given rule applies to. The asterisk (*) means that the instruction applies to all indexing bots on the internet.
• Disallow: Indicates a specific path (subpage or folder address) that should be excluded from scanning.

How do you add an entry to the robots.txt file?

Ads4media → cms4media portals → Edit → Settings → Rules for search engines - robots.txt file

In the ADS4media panel, go to the [cms4media portals] module. Find the portal in the list and go to its edit page. 

Find the Settings tab. There you will find the Rules for search engines - robots.txt file field.

Enter the rule and click Save.

The ads.txt file (short for Authorized Digital Sellers) is a list of trusted advertising partners. It is a simple text document added to your website in which you clearly declare which companies (e.g. Google) are authorized to sell ad space on it. 

Thanks to it, you protect yourself against online fraudsters who could impersonate your website, and advertisers can be sure that they are buying ad space directly from you. 

How to add an ads.txt entry?

Ads4media → cms4media portals → Edit → ads.txt Ad codes

In the ADS4media panel, go to the [cms4media portals] module. Find the portal in the list and go to its edit page.

Find the ads.txt Ad codes tab and enter the code in the Ad codes field. You can add a note, for example with information about which advertiser the code comes from. The note will not be visible in the ads.txt file. It is auxiliary information for Ads4media users.

Click the Add new ad codes button to save the changes. 

Before signing it, ask one question: what will a machine be able to read afterward that it cannot read today?

For many publishers with open, crawlable article pages, the answer is: not much. Modern search and AI systems can already parse ordinary web pages. The real problem is not whether a model can enter the archive. It is whether the model — or a human reader — can identify what is in there, who reported it, when it happened and how one story connects to the next.

This is why the New York Post’s new AI product matters even to a local newspaper with no AI budget: it reveals that the quality of the answer begins with the quality of the archive.

What Hamilton actually does

On August 11, 2026, New York Post Media Group launched Hamilton inside the New York Post and California Post apps. Built with Google Cloud’s Gemini Enterprise Agent Platform, Hamilton combines conversational search, personalized news briefings, recommendations and notifications.

The conversational search is the part local publishers should study. According to A Media Operator’s reporting, it searches five years of the publisher’s archive at launch, answers questions in plain language and cites the Post articles behind each answer. Newly published stories can be ingested and made searchable within seconds, Axios reported.

The boundary is deliberately narrow. New York Post CTO Ariscielle Novicio described the rule given to the system this way: “If we don’t write about it, do not even try to answer.”

Hamilton does generate conversational answers and personalized digests, so it would be inaccurate to say the software writes nothing. What it does not do is original reporting or independent editorial judgment. The Post controls the source material and editorial boundaries; the model retrieves, organizes and presents journalism the newsroom already produced.

Even one of Hamilton’s least glamorous components makes the same point. The Post built a caching layer so thousands of similar reader questions do not create thousands of fresh, paid model calls. The visible product is AI. The work underneath it is information architecture, cost control and editorial discipline.

Readable is not the same as useful

There is an important qualifier to the claim that an archive is already AI-readable. Its pages must actually be accessible under the publisher’s chosen terms. A broken URL, an accidental noindex, a login wall or content rendered in a way a crawler cannot retrieve is still a locked door.

“Readable” also does not mean “free to train on,” “licensed for reuse” or “available to every model.” Access, licensing and technical readability are different questions, and publishers should decide each of them deliberately.

Once a system has legitimate access, however, another problem appears. What happens when almost everything has been filed under one category called “News”? What can it establish about provenance when the byline on thousands of stories says only “Staff”? What does it learn from a photograph with no caption, no alt text and a filename such as IMG_4827.jpg?

At 4media, this is a recurring pattern during newspaper migrations. Inside the disorder may sit 20 or 30 years of the only continuous record of a town: elections, school boards, floods, business openings, obituaries, high-school championships and the people who shaped the community.

Usually, nobody made one disastrous decision. The disorder accumulated one deadline at a time. A broad category was faster than choosing the right one. “Staff” was easier than maintaining author accounts. Captions and tags felt optional because the article still looked fine when it went live. Two decades later, those shortcuts have become the architecture of the archive.

An AI interface does not repair that history by appearing above it. It exposes the condition the archive was already in.

Google is not asking for a second, AI-only website

Google’s current guidance for generative AI features in Search is unusually direct: the foundations of visibility in AI search are still the foundations of SEO. Google recommends valuable, non-commodity content, a clear technical structure and pages its systems can crawl and understand. It says there is no special schema.org markup required for generative AI search and advises publishers to evaluate supposed AEO or GEO services against the same established SEO principles.

That matters because “AI readiness” is already becoming a label that can be attached to almost any service. A new text file, a block of machine-generated FAQs or a promise to “chunk” every article for an LLM may sound advanced without improving the underlying journalism at all. Our analysis of whether AI systems actually use llms.txt found the same gap between adoption and evidence.

Seven fields that decide what the AI finds

The durable work is less fashionable:

1. Specific categories and consistent tags. “News” is a container, not a useful description. Geography, institution, beat and recurring topic labels give retrieval systems meaningful ways to narrow a search.
2. Named authors with real profile pages. Google’s people-first guidance encourages clear bylines and information about who created the content. E-E-A-T is not a single ranking factor or score. It is Google’s framework for thinking about experience, expertise, authoritativeness and — most importantly — trust. CMS4media’s Authors module connects a byline to a profile, biography and the writer’s other work.
3. Reliable publication and modification dates. An AI assistant trying to reconstruct a mayoral race or storm response needs to understand the sequence of events. Silent date changes and missing update notes turn a timeline into guesswork.
4. Captions and descriptive alt text. A photograph is part of the reporting. A vision model may be able to guess what an old image shows, but it should not have to guess who is pictured, where the photo was taken or why it matters. Captions preserve editorial context; alt text makes the page more accessible.
5. Stable URLs, canonical signals and honest redirects. A citation is only valuable if it still resolves to the story it names. Our guide to redirects and HTTP status codes for news archives explains why sending every retired article to the homepage creates confusion instead of preserving value.
6. Article-level structured data. Google says Article and NewsArticle markup can clarify a story’s headline, author, images and publication dates. It is not a magic ranking switch and not special AI markup. It is a consistent description of what is already on the page.
7. Source relationships and internal links. A developing story is rarely one URL. Links between the first report, public records, corrections, follow-ups and analysis give both readers and retrieval systems the surrounding record.

None of these fields can turn thin reporting into authoritative journalism. They do something more practical: they stop authoritative journalism from disappearing inside its own database.

The archive should work before the chatbot arrives

Hamilton is an audience product, not merely an archive demo. The New York Post team says it is moving beyond raw pageviews to measures such as visit frequency, session depth, retention, conversion and reader lifetime value. That is the right test: not whether an AI answer looks futuristic, but whether it helps readers build a deeper relationship with the publication.

A local publisher can start extracting the same kind of value without building a conversational assistant. Old reporting should appear beside current reporting. A reader following a new zoning dispute should be able to reach the original development proposal, earlier council votes and the reporting on the officials involved.

The Related Articles module in CMS4media, for example, can resurface stories by category or tag. That helps a human reader today and rewards the same classification work a future retrieval system will depend on tomorrow.

The migration is another decisive moment. In a recent CMS4media project, Hartmann Media Consulting moved six local news portals. On its larger sites, the work included years of articles, photographs and metadata, followed by redirect cleanup, indexing checks and Search Console monitoring. An independent audit cited in our migration case study found that organic search traffic held steady while Google Discover grew by close to 41 percent and Google News by 14.9 percent.

This is what responsible archive work looks like: preserve what has value, map the old URLs, expose technical problems and measure what happens after launch.

No responsible CMS can reconstruct a missing byline or invent a reliable caption for a photograph whose context has been lost. What it can do is preserve the information that still exists and make better habits easier from this point forward.

CMS4media’s AI Assistant, for example, can help editors fill repetitive fields such as titles, introductions and SEO descriptions from the reporting they have already entered. It can reduce the daily friction that caused metadata to be skipped in the first place. The newsroom still reviews the result. AI should help maintain editorial structure, not substitute for it.

Ask vendors three questions, not one

Soon every CMS vendor will say it has AI. Most will be telling the truth. The label will therefore tell publishers almost nothing.

Ask instead:

1. What have you already built and migrated? Look for live publishers, intact archives, working author records, stable URLs and measurable post-migration results.
2. What are you building now? A roadmap should solve problems editors and publishers can recognize, not simply follow the vocabulary of the latest keynote.
3. How is the next roadmap decided? At 4media, regular conversations with publishers and patterns in real support requests shape what we build next. That process matters more than the number of AI logos on a sales slide.

Your archive may need cleanup. It may need a careful migration. It may need better taxonomy, author records, captions, redirects and structured data. Those are real projects, and some require serious work.

What it probably does not need is an expensive costume called “AI readiness.”

The New York Post did not make five years of journalism valuable by placing Hamilton in front of it. Hamilton made the existing value easier for readers to reach. For a local publisher, the order of operations should be the same: protect the reporting, organize the record and make every source easy to retrieve. Then decide which new door you want to build into it.

Metadata will never be the most exciting line in a press release. It still runs the show.

But what happens when a new resident moves to town and asks ChatGPT or Google’s AI Overview, "Who is the most reliable plumber near me?"

If the AI can’t verify your existence across multiple trusted sources, to the machine, you simply do not exist. You will not be recommended.

The digital foundation of many local business organizations is showing serious cracks. Anecdotal evidence and industry observations suggest that across various states, local chamber directories often suffer from broken links, outdated platforms, or missing websites altogether. When a directory listing points to a defunct address or an unverified subfolder, it represents a catastrophic failure of local verification in the AI era.

The New Math of AI Local Search

When a large language model (LLM) decides whether a local business is real and worth recommending, it doesn't just read your website. It looks for consensus across sources it has been trained to trust. The big three are typically:

1. Google Business Profile
2. The local newspaper
3. The local Chamber of Commerce directory

Chamber listings rank among the highest authority local citations available. They are a definitive signal of legitimacy—one of the few that an automated SEO service cannot fake for you. This aligns perfectly with Google's emphasis on E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness). A verifiable link from a recognized local authority is a powerful trust signal.

The stakes are incredibly high. Data from SOCi, a leading local visibility platform, reveals a harsh reality: currently, only 1 in 80 local businesses ever get named when someone asks an AI for a recommendation.

With an estimated 80% of U.S. consumers searching for local businesses online every week, failing to appear in AI-driven discovery is a massive lost opportunity. AI visibility is exponentially harder to win than traditional map pack rankings, and it demands flawless consistency across trusted local data sources.

Verification Infrastructure, Not Just a Perk

This paradigm shift means a Chamber directory is no longer merely a "membership perk." It is critical verification infrastructure for every business in the county.

If a Chamber’s listing is a static PDF, or if the Chamber's own web address resolves to a broken link or an unverified free blog page, it is failing to vouch for its members at the precise moment a machine is deciding whether they exist.

This lack of digital discoverability couldn't come at a worse time for some regions. Southern Nevada, for example, recently experienced some of its slowest visitation periods and casino revenue dips since the pandemic recovery began in 2021. When foot traffic is down and macro-economic factors compress margins, local businesses rely even more heavily on digital discovery to capture remaining demand. It is a very bad year to be hard to find.

Building What the Machines Trust

The solution lies in robust, localized digital platforms that prioritize structured data, first-party control, and unshakeable authority.

Just as local newsrooms must build direct relationships with their audience to survive the AI transition (as discussed in our recent piece, Google’s "Velvet Rope": What New Search Profile Thresholds Mean for Local Publishers), community organizations must build infrastructure that AI engines can seamlessly read and trust.

This is why modern publishing systems are adapting. The technology required to power a trustworthy local news site is the exact same technology required to power a trustworthy community directory.

Platforms like 4community (an adaptation of the CMS4media infrastructure designed for publishers) are built specifically to solve this problem. Used by organizations like the White Pine Chamber of Commerce in Ely, Nevada, this system provides:

• Structured, verifiable directories: Ensuring every member business is presented in a way that AI models recognize and validate.
• High-authority domain structures: Moving away from third-party subfolders to owned, authoritative domains.
• Integrated local ecosystems: Connecting local news, business directories, and community classifieds into a single, trusted web of local information.

The Ultimate Test

The test for local legitimacy is no longer a manual search. It is a prompt.

If you run a local business organization, try it yourself. Ask ChatGPT, Claude, or Google’s AI: "Who should I call in [Your Town] for a plumber, an accountant, or an insurance agent?"

If your dues-paying members are not in the answer, your digital infrastructure is failing them. In the platform era, authority isn't just about who you know in town; it's about whether the machines can verify that you know them.

Here's the problem: given how search actually works today, that instinct backfires. Search engines read a blanket redirect-everything-to-homepage move as a technical error, not a save. The result isn't preserved rankings — it's lost rankings and wasted crawl budget.

This piece walks through the HTTP status codes that matter for a newsroom, one at a time, with the kind of real-world scenarios editorial teams run into constantly. You'll see what separates a permanent redirect from a temporary one, and when it's actually better to just tell visitors — and search engines — that a page is gone for good.

We'll also get into how AI systems handle these same codes, because that's turned into a different game entirely.

Status 200 (OK): The State Everything Should Be In

Let's start with the basics — status code 200.

It's the simplest possible message: everything worked exactly as it should. The server tells the browser and the crawler, "here you go, the page is ready."

This is the state you want for every URL on your site.

Articles, the homepage, category pages — all of it should return this code. No technical friction. The reader sees the content instantly, and the crawler indexes it without any obstacles.

Take a simple example: someone clicks a link to yesterday's game recap. The page loads normally, the content is there, the response code is 200. Nothing else needs to happen — and that's exactly the point. Traffic flows without friction because there's no intermediate step where anything could go wrong.

Status 301: The Classic That Preserves Your Link Equity

Now for status 301 — probably the best-known status code in SEO.

It's a clear statement: this content has moved to a new address, permanently. Used correctly, it's a genuinely powerful tool.

When you serve a 301, you're giving the search engine a specific instruction: forget the old link, replace it with the new one in the results. A crawler that had previously indexed the old article URL starts treating the new URL as the true home for that content — and that's the version that shows up in search results going forward.

But that's not all a 301 does. It also carries forward the SEO value that old URL built up over time. Picture an article from five years ago that dozens of other sites have linked to — blogs, other newsrooms, industry forums. Each of those links functions as a kind of trust vote in Google's eyes. A 301 makes sure those votes don't disappear along with the old URL — they carry over to the new one.

That's why 301s show up in a few specific situations. First: restructuring your URL scheme, say when a site moves from a clunky, technical URL format to something cleaner and more SEO-friendly. Second: fixing a typo in a headline that forced a change to the URL itself. Third: migrating an entire publication to a new domain, where every old URL needs a matching address on the new site. Fourth: merging two short, archival stories about the same event into one more complete article — you'd redirect both old URLs to that new, combined piece.

That power is exactly why so many damaging myths have built up around the 301 — and the biggest one is the idea that a 301 works like a universal patch: slap it on any dead URL and the problem disappears. In practice, a 301 only carries link equity forward when the new page genuinely continues the topic of the old one. Redirecting a deleted sports story to a homepage full of political news doesn't save anything — which brings us to the mistake covered next.

Why Redirecting a Deleted Story to the Homepage Is a Mistake

Newsrooms are terrified of blank pages and 404 errors.

That fear is where a bad habit comes from: delete an article from the archive, and the system automatically bounces the reader to the homepage. The logic seems sound — preserve the traffic, hold onto the link equity.

To Google, though, this is completely illogical behavior.

Picture someone searching for results from a local election a decade ago. They click the link and land instead on today's homepage with the weather forecast. Something feels off immediately.

Google recognizes this pattern easily. When it sees that the destination content has nothing to do with the original, it treats the redirect as deceptive — and triggers a defensive response.

The industry term for this is a soft 404 — because the page technically returns a 200 status ("OK"), but the content isn't actually what the user asked for.

The result? All the link equity you were trying to save simply doesn't transfer. It gets stuck halfway there.

The rule is simple: a 301 only makes sense when the new URL genuinely answers the same question the old one did.

Redirecting an old sports story to a new sports story on a similar topic — sure. Redirecting that same story to today's general news homepage — no.

Status 302, 307, and 308: Temporary Redirects and the Password Problem

Sometimes you need to route traffic elsewhere, but only briefly — during an outage in one section, or while testing a new layout.

For years, 302 was the tool for this: a temporary move. The message to the crawler was, "don't forget the old URL, it's coming back soon."

But it had a real technical flaw, one that was especially painful for subscription sites.

When a reader logged into their paid account and the system quietly ran a 302 redirect behind the scenes, older browsers could actually lose the submitted form data along the way. Take a concrete case: a reader fills out the login form, hits "sign in," and the system redirects them back to the article via a 302. Somewhere in that handoff, the login and password data could get dropped. The result: the reader has to type everything again, or just sees a login error — frustrating, especially for paying subscribers.

That's why newer standards emerged: 307 (temporary) and 308 (permanent). They behave the same way 302 and 301 do from an SEO standpoint, but they have one important advantage.

They never lose submitted data. Period.

When a reader submits a password through a payment gateway, 307 guarantees it reaches the new address safely, with no risk of anything getting lost along the way. Google trusts these codes fully, so it's worth making sure your engineering team uses them anywhere data is being submitted — logins, payments, forms.

Status 404: An Empty Page Isn't a Disaster

The internet's most famous status message: 404.

It simply means the server couldn't find the requested page. A persistent myth in marketing circles is that every 404 is a nail in the coffin for your Google rankings.

That's one of SEO's more stubborn myths.

The web is alive — articles get published and eventually disappear, and dead links are a completely normal part of running any large site. Google understands this and doesn't penalize you for cleaning up your archive.

There is a catch, though, and it matters for large publishers in particular.

When a crawler hits a 404, it assumes this might be temporary — maybe an admin is about to fix the link. Take an example: a site removes an old story about a water-main break from three years ago. The crawler hits a 404 and, instead of considering the matter closed, comes back to check that same empty page month after month — just in case the content returns.

That's crawl budget being wasted on something the crawler could be spending on more useful work, like indexing today's articles.

If your site is dealing with an indexing problem more serious than the occasional stray 404 — say, Google isn't seeing your site at all — that usually points to a deeper crawling or indexing issue worth auditing separately.

Status 410 (Gone): The Tool That's Easy to Overlook

There's a second code for the exact same situation as a 404 — but with a very different meaning attached.

A 410 differs from a 404 in one key way. It doesn't say "I couldn't find this." It says, plainly, "this was intentionally removed, and it isn't coming back."

That's a very specific signal to anything crawling your site.

When a crawler hits a long-deleted story about a local traffic accident and gets a 410 instead of a 404, it registers that decision and stops coming back. If the content is gone permanently, there's no reason to check again next week, next month, or next year.

That's a real benefit for the entire site.

The crawl budget a search engine spends scanning your site stops getting wasted on dead URLs. Instead of circling back to old, empty pages, crawlers get to your five-minute-old articles that much faster.

For news publishers, this often determines whether you show up in Google News at all.

Don't Forget the Human Who Actually Sees This Page

There's one thing that's easy to overlook when you're building out a page for deleted content.

Returning the right error code for the crawler is one thing. What that page looks like to an actual reader is a completely different problem. Nobody wants to land on a bare white page with a technical error message — that's the fastest way to lose someone for good.

It's worth building a dedicated, friendly view for these cases.

Something on-brand, with a simple message: "this story has been moved to our archive because it's no longer current." Add a search bar.

Below that, surface a few of today's best stories.

That way, Google gets the signal to clear the link from its index, and the reader stays on your site instead of bouncing.

Status 451 and 418: The Law, and an Industry Joke

The HTTP vocabulary also has codes for some very specific situations.

One is 451 — it sounds grim, and that's intentional. The name is a nod to "Fahrenheit 451," the novel about book burning. This code shows up when a publisher has to take down an article for legal reasons.

A court order, a law enforcement takedown request, a copyright claim, state censorship — in any of these cases, independent outlets generally prefer to show a 451 rather than quietly delete the piece. It's a clear signal that the removal came from outside pressure, not an editorial decision.

For organizations tracking press freedom, that distinction matters.

A more amusing case is status 418 — a pure industry joke.

It was created more than twenty years ago, as an April Fools' gag. It literally means: "I'm a teapot, I can't brew you coffee." It was meant as a harmless joke among developers, with no real-world use.

Today, though, it occasionally gets a genuinely practical application.

News sites regularly fight bots that scrape content en masse. Picture a script hitting a server hundreds of times a minute, copying article after article into its own database. When a defense system catches that pattern, it can deliberately return a 418 instead of the actual page content.

The confused script often just chokes, unsure how to interpret the response.

Redirect Chains: Getting Lost Inside Your Own Site

Now that you know what the individual codes mean, it's worth looking at what happens when they stack up into long chains.

A publication grows over years — an SSL certificate gets added, section names change, the "www" prefix disappears. A link from 2011 jumps to a 2015 version, then 2018, and finally lands on the current URL.

That's a redirect chain — and it's the enemy of any optimization effort.

Every hop in that chain costs time. On a fast office connection you barely notice it, but a reader on a weak mobile signal sees a blank white screen for several seconds. Some of them will just close the page before it loads.

Google treats these chains just as unfavorably.

Per Google's own documentation: past ten hops, the crawler gives up. It stops following the chain, and whatever link equity you were trying to preserve evaporates along the way. That's why your technical team should keep the structure flat — old link straight to the final URL, no historical layovers in between.

Are All Redirect Chains Bad? A Few Exceptions

There are situations where a short chain makes business sense and doesn't hurt anyone.

The first is affiliate links. When you recommend a product, you route the reader through a server that logs the commission before they land in the retailer's cart. That's a deliberate, revenue-generating redirect, and search engines have no issue with it.

The second is access verification on subscription sites.

To check whether a reader has a paid subscription covering a given article, the system may need to briefly route them through a separate authentication service before sending them back to the article.

As long as that process is short and built on safe codes — like the 307 mentioned earlier — search engines don't see any red flags.

AI and HTTP Status Codes: A Different Set of Rules

Finally, it's worth looking at what publishers are up against beyond classic Google search these days.

The traffic numbers coming out of 2026 make that clear. Multiple outlets have reported steep, sustained drops in Google referral traffic since AI Overviews rolled out — enough that companies including USA Today's parent and Reddit have publicly discussed blocking Google's crawler entirely rather than keep feeding a system that sends them less and less traffic in return. Industry groups tracking member sites have documented year-over-year referral declines across both news and entertainment publishers, and the debate has moved well past "is this happening" into boardroom conversations about whether staying visible in Google is even worth the cost anymore.

That backdrop matters here because language models and generative AI search engines pull knowledge from news sites to serve users ready-made summaries — and AI bots handle redirects noticeably worse than Google's crawler does.

Back to the same example: a deleted story redirected to the homepage.

A classic search crawler will quickly spot the trick and disregard that URL. An AI bot tends to be far more credulous — it'll follow an old link to a deleted TV review, land on today's homepage featuring a political scandal interview, and simply merge the two as if they were connected.

The result? The model generates nonsense for the user, citing your newsroom as the source.

That's a fast way to burn reader trust in your brand — and at a moment when every newsroom is already fighting to hold onto whatever traffic and credibility AI systems leave them, that risk is harder to shrug off than it used to be. Therefore, using 410 honestly, wherever content is genuinely gone for good, is more than a technical best practice today — it's protection against misinformation that can spread across the web with your name attached to it.

A Few Questions Before You Hit "Redirect"

A large archive means constant decisions, and every one of them has a cost — in traffic, in team time, or both.

Knowing the 300 and 400 status code families isn't just an IT department concern. It's a strategic tool for anyone responsible for how a site shows up in search.

Before ordering another mass redirect of old topic sections, it's worth asking your team a few questions.

Does the new page genuinely answer the same question the old headline did?

Is a ten-year-old story about a long-forgotten local event still delivering any real value?

Would removing that content from the index via 410 free up crawl budget for fresher, current publishing?

If the answers point toward cleaning up the archive — do it. Build a proper error page, cut the dead links loose, and let the crawlers focus on what's actually current.

Imagine a local newspaper that has dutifully covered city hall for 70 years. It employs six dedicated reporters, maintains a robust historical archive, and is the lifeblood of its community’s civic information. Yet, under Google’s latest framework, it might fail the test to be recognized with a dedicated Search Profile.

Meanwhile, a food review account launched just last year can breeze past the velvet rope simply by holding a large Instagram following.

That is how Google Search Profiles work in the United States today. Marketed to creators, brands, and publishers, these profiles offer a substantial reward: a customizable space connected directly to Search and Discover. Qualifying entities get the chance to shape a Knowledge Panel, pin posts, add direct links, and let readers follow their source right from the search results page.

But the gatekeeper isn't journalistic rigor or local impact; it's social media followers.

The Follower-Weighted Era

Historically, the fundamental unit of the web was the page. Today, the platform era’s new unit is the audience account.

Initially, Google set the entry bar incredibly high: 100,000 followers on YouTube, Instagram, or X (formerly Twitter), or 300,000 on TikTok. While Google recently (August 2026) lowered these thresholds slightly for some platforms—dropping the requirement to 35,000 subscribers for YouTube, Instagram, and X (though keeping TikTok at 100,000)—the underlying philosophy remains unchanged.

This is follower-weighted legitimacy.

While Google explicitly states that creating a profile does not directly change search rankings, the implications for visibility are real. Users who follow a source via their Search Profile may see more of that creator's content surfaced in Google Discover.

For local publishers, this trend is a stark reminder: platforms evaluate worth through the lens of global scale, not local necessity. Google’s threshold tells us where its product is going. It should not tell a local newsroom what it is worth.

Why the Timing Matters (And What It Means for AI)

The push toward audience accounts comes at a critical juncture. As we've tracked in our recent coverage, including The Messenger Becomes the Author and Google Search Live and the End of Drive-By Traffic, publishers are facing an existential question: How much of our reporting should we keep feeding into AI search ecosystems as traditional referral traffic continues to fall?

Google is elevating social accounts to first-class properties inside Search Console, accommodating creators who might not even own a traditional website. This shift underscores a broader industry pivot: reliance on third-party platform traffic is becoming a losing game for hyper-local entities.

The Strategy for Local Newsrooms: Audience Sovereignty

Local publishers cannot afford to play by "national creator math." Instead, they need a two-pronged approach:

1. Claim Every Available Surface If your newsroom meets the new 35,000 follower threshold on a platform like X or Instagram, claim your Google Search Profile immediately. Optimize it, pin your best investigative work, and use it as a funnel. Play the game where you can.

2. Build What Platforms Cannot Revoke The ultimate defense against shifting algorithm thresholds is Audience Sovereignty. Local publishers must relentlessly build direct relationships with their readers.

This means prioritizing:

• Direct website readership
• First-party data collection
• Paid subscriptions and paywalls
• Owned newsletters

Platforms can change the rules of the velvet rope overnight. They cannot take away a loyal subscriber who visits your homepage directly or opens your morning email.

How CMS4media Supports True Independence

Building a sustainable local news model requires the right infrastructure—one that doesn't rely solely on algorithmic goodwill. This is where moving beyond a basic website to a comprehensive publisher platform becomes vital.

CMS4media is designed specifically to help publishers build that unshakeable direct audience.

• First-Party Revenue Streams: Rather than waiting for platform payouts, CMS4media allows publishers to easily implement Paywalls & Subscriptions, establishing a direct financial relationship with readers.
• Engagement Tools: Modules like the Content Hub and Quizzes keep readers on your site longer, turning casual visitors into habitual readers.
• Sovereign Monetization: With built-in features for a Classifieds marketplace, Business directory listings, and seamless integration with the Ads4media network, local newsrooms can diversify their revenue away from programmatic display ads and build local business partnerships.
• AI Efficiency: The built-in AI Assistant helps resource-strapped newsrooms optimize content for SEO and generate related materials quickly, ensuring that when you do compete for search traffic, your content is perfectly primed.

The Bottom Line

Google’s evolving Search Profiles are a fascinating tool for creators with mass appeal, but they expose a blind spot when it comes to the vital role of local journalism. A local publisher's worth is defined by the community it serves, not an arbitrary follower count.

By focusing on direct readership and utilizing tools designed for independent media growth, local newsrooms can ensure they remain the indispensable town crier, regardless of who is allowed past Google's velvet rope.

But as the standard has grown in popularity, the claims about it have gotten bolder. Some call llms.txt "robots.txt for AI." Others suggest it improves your odds of being cited by ChatGPT or Perplexity.

The problem is that opinions are multiplying faster than evidence.

Server log analyses suggest that most published files are barely ever fetched, and Google hasn't confirmed that llms.txt plays any meaningful role in its systems.

That gap between hype and evidence is exactly what makes llms.txt one of the more interesting topics at the intersection of SEO and AI right now. So instead of jumping straight to "does llms.txt work," it's worth looking at what we actually know. 

Where did the standard come from? What does its author actually propose? How are AI model providers responding to it? And does the available data really back up the narrative that's built up around it?

Where Did the Idea for llms.txt Come From?

The standard wasn't created by Google or OpenAI, and its origins have nothing to do with search engines.

It comes from Jeremy Howard, co-founder of Answer.AI and creator of fast.ai, who published the first version of the spec in September 2024.

Howard wasn't trying to invent a new ranking signal or an alternative to robots.txt. His starting point was a much more practical problem: modern websites are designed primarily for humans and browsers, while language models need the simplest possible path to the actual content that matters.

In the spec's description, Howard points out that a single web page often contains a huge amount of material that adds zero value from a model's perspective. Menus, footers, ads, JavaScript, banners, and heavy navigation all inflate the number of tokens needed to process a document, making it harder to reach the actual knowledge inside.

That observation is what led to llms.txt. The original goal wasn't SEO — it was making it easier for language models to find the most important content on a site.

The Problem Jeremy Howard Was Trying to Solve

The project's premise can be summed up in one sentence: since language models read the web differently than people do, maybe they need a different way to navigate it too.

Howard's proposal: a site owner writes a simple text file listing the site's most important resources, each with a short description. Instead of parsing the entire structure of a site, an AI agent gets a ready-made list of where to start.

That distinction matters. The spec doesn't propose replacing a site's existing structure or building a new indexing mechanism. The file is only meant to point to whichever documents the author considers most valuable.

You could say llms.txt proposes an additional layer of communication between a site owner and AI systems. Howard himself describes it as a guide to a site — not a new way of publishing content.

Is llms.txt an Official Web Standard?

Not yet.

This is one of the most commonly overlooked distinctions in discussions about llms.txt. Despite its growing popularity, the project hasn't been adopted by standards bodies like the W3C or IETF. That means it has no official status as a web specification, and implementing it is entirely voluntary.

That has real consequences. No AI model provider is obligated to read the file. No site owner is obligated to publish it. The whole project's success depends entirely on whether both sides find it useful.

That's what separates llms.txt from standards that eventually became a fixed part of internet infrastructure. Right now, llms.txt should be treated as an open proposal from the AI community — not a binding web standard.

What Does the llms.txt Spec Actually Look Like?

Simplicity was one of the project's core goals.

Per the spec, the file should live at /llms.txt in a domain's root directory. Its content is written in Markdown, not XML or JSON — a deliberate choice. Howard argues Markdown is one of the formats language models recognize best, since so much technical documentation is already written in it.

At minimum, the document includes a project name, a short site description, and sections grouping the most important resources with brief descriptions and Markdown-formatted links.

What the spec recommends against is just as telling. Howard doesn't suggest copying an entire sitemap or listing hundreds of URLs. The idea is the opposite: a short, hand-picked set of the most important materials. llms.txt is meant to read like a carefully curated table of contents, not an auto-generated list of every page on a site.

Why Isn't This Just Another sitemap.xml?

At first glance the two look similar — both contain links to pages. That's about where the similarities end.

sitemap.xml helps search engines discover URLs to crawl and index. In practice it often lists thousands of pages and is generated automatically by a CMS or SEO tool.

llms.txt does something completely different. It doesn't try to describe an entire site. It's meant to point to a small number of materials that, in the author's judgment, best represent the knowledge the site contains.

The difference isn't just technical — it's philosophical. A sitemap answers "what's on this site?" llms.txt tries to answer "where should you start?"

Actual Adoption of llms.txt: Industry Enthusiasm Is One Thing, AI Usage Is Another

A rising number of implementations doesn't mean a rising amount of usage

The number of domains publishing llms.txt is growing fast. In the roughly 15 months since the spec was published, the file has shown up on hundreds of thousands of sites, and popular SEO tools and documentation platforms have started offering it as a built-in generator.

That doesn't mean language models are actually using it. Site owners implementing the file and AI systems actually reading it are two separate questions — and that gap is really the whole debate.

BuiltWith shows the scale of adoption

BuiltWith's data confirms rapid growth in the number of sites with an llms.txt file. But that only shows that the technology is present — not whether AI models are using it. SEO has a long history of tactics that got adopted en masse mainly because they were bundled into CMS platforms and plugins. A rising implementation count isn't, by itself, evidence that the standard matters in practice.

Ahrefs checked server logs

Ahrefs ran a much more revealing analysis, based on HTTP logs from more than 137,000 domains. Instead of counting implementations, they checked whether llms.txt files were actually being fetched.

The results were sobering: roughly 97% of the files received zero HTTP requests during the study period. That doesn't prove AI models never use them, but it does show that actual usage of the standard is far from widespread today. What's more, a large share of the traffic that was recorded came from monitoring tools and technology scanners — not from systems generating answers for real users. A file being downloaded doesn't necessarily mean a language model made use of it.

OtterlyAI reaches similar conclusions

OtterlyAI published comparable findings. Looking at more than 60,000 AI bot visits, requests for llms.txt accounted for only about 0.1% of that traffic.

Two independent studies point to the same conclusion: llms.txt implementations are growing fast, but there's currently little evidence of widespread use by AI systems.

What Does Google Say About It?

Google's position is far more cautious than the industry conversation around it. John Mueller has said Google's systems can analyze page structure on their own and don't need an extra file to understand a page's content. Google hasn't communicated any ranking benefit tied to implementing llms.txt.

Google has also pointed out that the file's content is written by the site owner — meaning it's a self-reported signal. That kind of information is easy to manipulate, so search engines tend to weight signals they observe independently more heavily than claims made by page authors themselves.

Right now, there's no basis for treating llms.txt as equivalent to robots.txt or sitemap.xml, or as a ranking factor on Google.

If Models Can Already Read HTML, Why Do They Need Another File?

This is the skeptics' central argument

Discussions about llms.txt often boil down to one assumption: if language models get a "better map" of a site, they should produce better answers.

The problem is that modern models already have very sophisticated tools for parsing documents. Both search systems and language models can distinguish main content from navigation, ads, footers, and interface elements — that's not some new AI-specific skill; classic search engines have used similar techniques for years.

If a model can already identify the right part of a document on its own, it's fair to ask what additional value a hand-curated guide really adds. This is precisely the sticking point that keeps coming up between supporters and skeptics of the standard: the better models get at parsing pages, the harder it becomes to justify an extra descriptive layer on top.

Why the Distinction Between a Training Crawler and a RAG System Matters

Not every AI system visits a site for the same reason

A training crawler works much like a classic search-indexing bot: its job is to gather a huge volume of documents that will later be used to train a model. In that scenario, what matters most is the ability to pull large datasets and clean them up afterward.

Retrieval-Augmented Generation (RAG) systems work completely differently. They're not trying to build a full copy of the internet. Their job is to find the right documents at the moment a user asks a specific question.

This distinction changes how you should think about llms.txt — whether the standard is useful at all depends heavily on which stage of a model's workflow you're talking about.

In RAG systems, speed is what matters

Modern AI search tools generally work in stages: first they find potentially relevant documents, then fetch their content, then hand that content to the language model that generates the actual answer.

That raises an obvious question: if a system already knows the answer lives at a specific URL, why would it make an extra HTTP request just to fetch llms.txt, read a list of recommended materials, and then go right back to the page it already found?

This is where the biggest doubts about the standard's practical usefulness come from. Every extra step adds latency and adds another operation for the system to perform. In an environment where milliseconds matter, that's not a trivial argument. If an AI search tool already knows the right document, an extra guide file may simply be redundant.

There are scenarios where llms.txt makes more sense — developer tools are the best example

The fact that Google is cautious doesn't mean the whole standard is worthless. Quite the opposite — research suggests the strongest interest in llms.txt isn't coming from classic search engines but from the developer-tools ecosystem.

Coding assistants, task-executing agents, and systems that parse technical documentation typically work with much smaller, more focused datasets than general web search does. In that context, a hand-curated guide to a documentation set really can cut down the time it takes to find the right information.

It's no coincidence that some of the earliest adopters of llms.txt were API documentation sites, developer libraries, and open-source projects — an environment that fits Howard's original premise far better than a news site or an e-commerce store does. The most compelling use cases for llms.txt today are in technical documentation, not broad-based SEO.

Separating the Standard's Potential From Its Current Role

Weighing the Ahrefs data, the OtterlyAI analysis, and Google's stated position together, it's hard to argue that llms.txt has become a widely used piece of AI infrastructure. It's just as hard to call it a pointless project.

The spec solves a specific, real problem its author described, and it has real applications in specific scenarios — especially where models are working with technical documentation or performing agentic tasks. That's not the same thing as confirming broad adoption by the AI search engines handling millions of queries a day.

That distinction is central to the whole debate. Instead of asking whether llms.txt "works," it's more useful to ask who it was designed for and in what environment. Only then can you fairly assess its actual potential.

The Case For and Against llms.txt: Why Is the SEO Industry Split?

The controversy around llms.txt isn't really about the spec itself — the format is simple and easy to implement. The disagreement is about whether it solves a real problem for today's language models.

You can think Howard's idea is logical and still doubt that today's AI systems actually need it. In practice, the question isn't whether llms.txt is a valid standard — it's whether models have any reason to use it.

Websites are built for people, not models

Supporters point out that modern web pages are full of elements that are just noise to a language model. llms.txt is meant to shorten the path to the most important material and flag priorities — not replace HTML or sitemap.xml. In this view, the standard acts as a guide, not a new publishing format.

Models are getting better at doing without it

Skeptics point out that modern models are increasingly good at recognizing document structure on their own and separating real content from navigation elements. In RAG-based systems, reading an extra file means another HTTP request and added processing cost. If a model has already found the right document, an extra guide may just be dead weight.

Both sides have a point

Current evidence suggests llms.txt may be worthwhile in specific use cases — technical documentation and agentic systems, for example — while there's no evidence it plays a meaningful role in mainstream AI search today. That makes it worth treating as a promising experiment, not a mandatory piece of technical SEO.

What Can We Actually Confirm?

The standard exists and is gaining traction quickly. There's no question llms.txt is an active, evolving project. Since the spec was published, more implementations, generators, and documentation-tool integrations have appeared, and adoption is climbing fast. The debate isn't about whether the standard exists — it's about how much it's actually being used.

There's no evidence it affects Google. Google hasn't confirmed that llms.txt influences rankings or AI Overviews. John Mueller has emphasized that Google's systems can analyze page structure on their own and don't need an extra file to understand what a page contains. There's currently no basis for treating llms.txt as a ranking factor.

What Should Be Considered a Myth?

"llms.txt works just like robots.txt."

This is one of the most repeated oversimplifications. Yes, both files live in a domain's root directory — but they serve completely different functions. robots.txt talks to crawlers and sets rules for what they can access. llms.txt doesn't block anything and has no effect on indexing; its only role is to flag the materials a site owner considers most important. The robots.txt comparison works fine as a loose metaphor, but technically the two solve entirely different problems.

"Implementing llms.txt improves SEO."

There's currently no data supporting this. No available study has shown an increase in organic visibility after implementing the standard, and Google hasn't indicated it treats llms.txt as a ranking signal. That could change down the road, but as of today, linking llms.txt to classic SEO outcomes goes beyond what the evidence actually shows.

"Without llms.txt, AI won't use your site's content."

This isn't supported either. Language models have been analyzing websites for years without llms.txt, relying on crawlers, their own indexes, and search mechanisms. When an AI system does draw on a site's content, it's almost always thanks to existing indexing infrastructure — not because it happened to find an llms.txt file. Not having llms.txt doesn't make a site invisible to AI systems.

What Should Publishers Take Away From This?

If you run a website, you can implement llms.txt — but right now there's no reason to expect it to move the needle. Putting the file together isn't hard and usually doesn't cost much. The issue is that there's currently no evidence tying its implementation to better Google visibility, more frequent AI citations, or more traffic.

That doesn't make llms.txt worthless. If you run an extensive documentation site, a knowledge base, or a resource-heavy expert content site, a curated guide like this could turn out to be useful down the road, especially as AI agents become more capable. For now, it's best treated as a low-cost experiment whose benefits haven't been proven yet.

For most publishers, the bigger wins will keep coming from tactics with a proven track record: producing genuinely valuable content, improving information architecture, using structured data well, and keeping your site fast. That's exactly where CMS 4Media earns its keep — with built-in SEO automation, structured data handled for you, and an AI assistant built into the editorial workflow, it takes care of the fundamentals that actually move the needle, so your team can spend less time on technical busywork and more time on the content itself.

• They use the AI Assistant module.
• They have specifically two language versions enabled simultaneously.

Tip: Detailed information on language configuration and instructions on how to activate additional language versions of your portal can be found in the following article:

Where Can You Find the Translation Feature?

Automatic translation works in most sections of your panel. You can use it in the following modules:

• Articles
• Authors
• E-editions
• Information ticker
• Photo galleries
• Multimedia
• Live coverages
• Boards
• Polls
• Plebiscites
• Events
• Obituaries
• Subpages
• Categories
• FAQ
• Paywall
• Classifieds
• Business Directory
• Terms
• Quizzes

How Does Language Selection Work?

The language you are translating from depends on the language version of the Admin Panel you are currently working in:

• Working in the English version of the panel: The button will trigger a translation from English to the second active language (e.g., Spanish).
• Working in the Spanish version of the panel: The button will trigger a translation from Spanish to English.

To change the language version of the admin panel itself, click the flag icon in the top right corner of the screen.

How the AI Translation module works.

The following description is based on the [Articles] module. The feature works exactly the same way in all other modules.

1. Prepare and save the text in your native language (e.g., in English).
2. Click the translation button (e.g., [English > Spanish]) in the top right corner of the form.
3. The AI Assistant will start working and display a message indicating that the translation is being prepared.

Once the process is complete, a green success confirmation message will appear at the top of the screen.

In addition, a flashing green dot will appear to indicate changes that have been made to the currently inactive tab of the second version of the post. It disappears when you switch to the aforementioned ES tab.

Important Rules for AI Assistant Operation

Preserving Formatting: The translated text retains its original layout. The AI copies the structure of headings and paragraphs, and correctly transfers links and images embedded within the content.

Content Protection (No Overwriting): The AI Assistant does not overwrite fields that you have already filled in manually. If you enter your own title in the Spanish (ES) tab before running the AI, the system will skip that field and leave it unchanged.

Empty Fields: The AI only translates fields that contain text. If you leave a field empty, the system will skip it. If the AI Assistant does not detect any content to translate on the page, it will display an appropriate warning message.

Image Metadata (Descriptions and Titles): The system also automatically translates the descriptions and titles of the main image and additional images, provided they were previously entered in the source language.

Translating Individual Images: If you are editing the metadata of a single, previously added image, you can find a dedicated translation button directly inside the edit panel for that specific image.

Key results at a glance

• 6 portals migrated to CMS4media, live since April 1, 2026
• Google Discover traffic up close to 41 percent month over month
• Google News traffic up 14.9 percent
• Organic search traffic retained through the migration, with no meaningful drop
• The Koronowo portal grew from roughly 90,000 to roughly 115,000 monthly users, and now stays consistently above 100,000
• Editorial team productive on the new system within about 5 days

Who this is about

Przemysław Hartmann runs Hartmann Media Consulting, the business formerly known as the Kujawy Media Group. The company has been on the market for about 12 years, and his flagship title, mojakruszwica.pl, just passed its 12th anniversary. He publishes across a cluster of local sites in the Kujawy region of Poland, including portals in Koronowo, Strzelno, and Janikowo, each serving a different small-town audience with different interests.

He did not start as a software person. He started as a journalist and editor, then grew into a publisher when it became clear the business had to be run as a business, not just a place to post news. That background matters, because it makes him exactly the kind of reader this case study is written for: a hands-on local publisher, not an IT department.

The starting point: two kinds of pain

Hartmann had been through several content management systems before CMS4media, and the history was not happy.

In the early years, his team ran home-grown, frankly amateur solutions. He remembers the "blood, sweat, and tears" of that first portal, a site that broke down often and left him and a partner to fix every problem alone, with no one to call.

His most recent provider solved the "build it yourself" problem but introduced a new one: advertising. Heavy, resource-intensive ads were getting between readers and the content. Readers complained. As he put it, people today expect straightforward access to a story, not a page that makes them wait while ads stack up. When a page takes several seconds to load, the bounce rate climbs, because almost no one is that patient anymore.

There was also a fear that keeps many independent publishers out of any network: the worry about losing independence and having someone interfere in the newsroom workflow. Looking back, he says that fear did not come true.

The part every publisher dreads: migration

Migration is the moment publishers lose sleep over, because it puts the archive and the search traffic at risk at the same time. Hartmann faced a real decision on each site: migrate the full database, or archive it and start clean. On some portals his team archived the old database. On the heavier sites, including Koronowo and Kruszwica, they migrated the entire archive, years of articles, photos, and metadata.

A migration at that scale is not a one-click event, and the honest version of this story includes that fact. After go-live, the portals were checked with a post-migration audit on the 4media side, and the work also included an independent SEO audit of the result. That review drives a prioritized cleanup of the usual long-tail migration issues, things like setting 301 redirects on the highest-traffic legacy URLs, retiring an old service worker left over from the previous setup, and resolving indexing notifications from Google, with follow-up checkpoints in Google Search Console to measure each fix. None of that is glamorous, but it is the difference between a migration that holds and one that quietly leaks traffic for months.

The aggregate result, which Hartmann read out himself on camera, was the part that mattered to him. Google Discover grew by close to 41 percent. Google News grew by 14.9 percent. Search traffic was effectively held in place rather than collapsing. Old broken links from the previous system stopped being a problem. (See roughly the 19:25 to 23:00 mark in the interview.)

He is also candid about the human side. Some traffic loss always happens at first, because readers have to get used to a new site, and reactions are mixed in the first weeks. April was a test month. By June he could see the numbers settling and then climbing. The Koronowo portal, his most representative site, moved from around 90,000 monthly users last summer to around 115,000, and now holds above 100,000. Crucially, that growth was not only local readers, which was a specific goal for a portal based in a very small town.

What changed in daily work

Speed came first. The heavy, intrusive ads are gone, and the site stays fast even when it is carrying a lot. Hartmann stress-tested this by pasting in HTML-heavy material, building an entertainment zone, even embedding a game. On his old setup that would have dragged the whole template to a crawl. On CMS4media it did not freeze or slow the page.

Then came the breadth of the platform, which he describes not as a place to add articles but as a full "media environment." A few modules stood out in his daily use:

• The Classifieds module finally gave his listings a place to be promoted, and reader interest in them took off in a way the old template never allowed.
• Quizzes and other engagement modules keep readers on the page and coming back.
• Photo galleries became a real tool rather than a dump of images. He tests different formats against his audience, including the hybrid gallery that shows a few photos and then prompts a click-through, and the scroll-based cascade gallery, trading off between time on page and page views.
• Related Articles let him link stories together and resurface older material from the archive, which keeps readers moving through the site.

For a publisher running six titles, the multi-site capability is the quiet hero. He can manage and share content across portals from one place. His early fear was that sharing the same content across sites would get him penalized by Google for duplicate content. It does not work that way here, because the system points the duplicates back to the source portal, so the main site is the one that benefits and SEO is not hurt. That means a story that fits two towns does not have to be written twice.

He is also testing the platform's AI assistant, which generates suggested titles, tags, and SEO-oriented material, and lets him set tone and format. For quick news, where typing tags by hand is a chore, it works instantly. He was wary of the common claim that AI tools hurt a portal's ranking, and his experience is that they speed up the work without that downside. Behind the assistant sit more than 70 prompts and a different model matched to each module, which is the reason it behaves sensibly rather than generically. (See roughly the 24:00 to 26:00 mark.)

The "free CMS" question

Plenty of small publishers start on a free CMS and never want to add a cost. Hartmann understands that instinct, and he pushes back on it from experience. After more than a decade online, his view is blunt: there is no good software that is truly free. Someone has to recover the cost, and with a free CMS they usually recover it by cluttering your pages with the very ads that drive readers away. In an era when publishers already have to fight social platforms for every visit, paying that hidden tax in ad clutter is, in his words, not worth it. He would rather pay a subscription for a stable, professional, flexible system and run a portal he is proud of. He moved six sites over knowing that meant real cost. (See roughly the 14:30 to 16:30 mark.)

On the revenue side, 4media supplies ads and sponsored articles, and the two sides split ad profit 50/50. What impressed him was the transparency: the ad report was clear and detailed rather than a black-box number in a Word document, he could see where the amounts came from and how engagement fed into them, and when he sent follow-up questions he got specific answers. The contract itself was negotiable, adjusted along the way to fit how he works.

The FREE Trap You Never Knew Existed

Onboarding and support

A powerful system raises a fair fear: will it be too complex to use. Hartmann's first reaction was exactly that, a large system and the worry of figuring it out. In practice it took about five days, maybe less, to feel at home, helped by four or five solid training sessions. His verdict is that the system is extensive but it makes sense, and that drag-and-drop content handling makes everyday publishing genuinely pleasant.

Support is the other half. He felt, especially during the migration, that the team was essentially at his disposal, to the point he worried he was leaning on them too much. When Google flagged indexing issues, he got help getting the archive indexed and available. When an ad-provider glitch surfaced content that should not have appeared, that got resolved too. Beyond day-to-day contact with the team, there are quarterly roadmap meetings where publishers can raise ideas for what gets built next.

What comes next

Part of why Hartmann recommends the platform is that the runway keeps going. He can cover live events, add audio and radio streams, and is preparing an e-edition, a PDF version of his printed newspaper sold behind a paywall, alongside plans to move his radio onto the portal so people can hear the local news there too. For a publisher who also prints, that combination of print, web, and audio under one roof is the point. As he sums up the platform's flexibility, "basically nothing limits us now."

Would he recommend it to other publishers?

Yes, without hesitation. After nearly three months he says it works the way it should and makes sense, the team has ideas for the future, and the experience extends beyond Poland, including publishers in the United States and the server-side solutions behind them. For any publisher, in print or online, looking for serious, professional tools, his advice is simple: it is worth a conversation, and worth testing. (See roughly the 28:00 mark.)

Frequently asked questions

Will I lose my SEO or search traffic if I migrate my news site? Migration always carries risk, and some short-term wobble is normal as readers adjust. In this case, organic search traffic was retained through the move, while Google Discover rose by close to 41 percent and Google News by 14.9 percent, supported by a post-migration audit and a prioritized cleanup of legacy redirects and indexing issues.

Is a paid news CMS worth it compared to a free one? This publisher's view, after more than a decade online, is that no capable software is truly free, and that free systems often recover their cost by loading your pages with intrusive ads that cost you readers. He chose to pay for a stable, flexible, professional platform instead.

Can I manage several local sites from one place without a duplicate-content penalty? Yes. The multi-site capability lets you publish and share content across portals from a single admin, and shared content points back to the source portal, so the main site benefits and search rankings are not harmed.

How long does it take to learn the system? For this team, about five days to feel comfortable, supported by four or five training sessions and ongoing support.

Does using the AI assistant hurt my SEO? In this publisher's experience, no. The assistant generates titles, tags, and SEO material and speeds up routine work, with a different model matched to each module rather than one generic engine.

If you do not publish in California, do not look away. Illinois, New York, and New Mexico already run versions of this idea, and more than a dozen states are drafting their own. This is a template, not a local curiosity.

The number is the headline. The architecture is the story. Read the bill closely and the intent is unmistakable. California is not subsidizing local news in general. It is paying for a particular kind of it: full-time, locally resident, independent, and paid like a profession.

Reading the Architecture

Four design choices reveal what the state is actually buying.

Refundable, not a deduction. Credits beyond an outlet's tax bill are paid out in cash, like a grant. A nonprofit with no tax liability still receives a check. The money reaches the newsrooms that need it most, not only the ones already turning a profit.

A 50-mile residency rule. Qualifying journalists have to live within 50 miles of the area they cover. That single clause quietly outlaws the remote-stringer model dressed up as local news.

A salary floor and real hours. Credits attach to full-time jobs of at least 30 hours a week paying no less than $35,000 a year. Part-time roles earn a smaller $7,500 credit. The state is drawing a line against gig-economy journalism and the side-hustle byline.

A 15% cap on political money. An outlet controlled by a PAC or a 501(c)(4), or drawing more than 15% of its revenue from them, does not qualify. Editorial independence is treated as a precondition, not a line in a marketing deck.

A Math Problem, Not an Audience Problem

The bet underneath the bill deserves to be stated plainly. Local news did not collapse because readers stopped caring. It collapsed because the unit economics of a reporter's salary broke roughly a decade ago and never recovered. Subscriptions, events, and the occasional grant could not close the gap on a full-time wage.

The scale of the damage is on the record. The United States has lost about 75% of its local journalists since 2002. California, for all its size and wealth, ranks 42nd of 50 states, with roughly six full-time local journalists per 100,000 residents, against a national average near 40 per 100,000 two decades ago.

AB 2222 aims straight at that bottleneck. It does not chase audience. It lowers the cost of keeping a reporter on payroll, which is the exact line item that broke.

The Catch

This is not money in the bank, and it is worth being honest about why. The bill is not law yet. It still has to clear the Senate, where budget hawks are pushing for tighter guardrails, and the program is estimated to cost the state around $22 million in its first year and roughly $50 million a year after that. California's tax agency has already flagged practical questions, including how to measure baseline headcount and how to stop outlets from gaming the credit. Critics also note that nonprofits, which already avoid most taxes, would receive what amounts to a second subsidy.

History argues for patience too. In 2024, a high-profile, Google-backed deal promised California newsrooms a large support package, and much of that money still has not reached individual publishers. Treat any policy dollar as pending, not banked.

Why This Matters Outside California

The credit itself will not reach a publisher in Ohio or Idaho. The precedent will. New York has committed $30 million a year to its version, and California's bill has now cleared its first chamber, which is exactly the kind of momentum that moves an idea from one statehouse to the next.

Two things follow for any publisher, in any state. First, watch your own legislature, because this is the most credible policy lever local news has gained in years. Second, understand what every one of these laws rewards: documented, qualifying, independent newsrooms. The money flows to outlets that can prove what they are.

The Half a Tax Credit Cannot Fix

A credit repairs one side of a reporter's unit economics, the cost. It does nothing about the other side, the revenue that reporter generates. A $20,000 credit buys time. It does not, on its own, build a business.

This is where we spend most of our time with publishers at 4media. The newsrooms that turn policy support into lasting stability share one trait: a platform that converts reporting into recurring revenue, through paid memberships, newsletters, and reliable ad delivery, so that every reporter the state helps fund also earns. That is the same shift toward owned, durable audiences we traced in Google Search Live and the End of Drive-By Traffic, and the same logic behind public money rewarding readiness in The $129 Million Scaffolding.

There is also a paperwork reality. To claim these credits, an outlet has to document full-time status, residency, salary, insurance, editorial independence, and, for digital outlets, that at least a third of its audience sits in-state. When we audit publishers, the most common gap is not editorial, it is data: many cannot produce a clean, defensible audience report, which is precisely what these credits require. Proving what you are takes systems, and aging software that cannot generate a reliable traffic number will leave money sitting on the table.

The Real Question

California will not fix local news with one bill, and the Community NEWS Act does not pretend to. But of every proposal in the past five years, this is the one aimed at the actual bottleneck: the broken math under a reporter's salary.

So the open question is not really whether the policy works. It is whether other states copy it, and whether your newsroom is built to claim the money when they do. California has decided a local reporter is worth $20,000. The reporters covering your community are worth more than that to you. Build the platform that can prove it before the check is on offer, not after.

(Source: AB 2222, Community Newsroom Employment and Workforce Sustainability Act, California Legislature)

Community event coverage is often treated as disposable content. Editors publish a quick announcement about this weekend's farmers market, a summer concert series, or a county fair, and once the event is over, the page quietly fades into Google's archive.

Meanwhile, thousands of people continue searching for phrases like "things to do this weekend," "free events near me," "family activities in Phoenix," or "live music in Nashville tonight." These highly specific searches may not generate blockbuster traffic individually, but together they represent one of the most valuable long-tail traffic sources available to local publishers.

An event calendar shouldn't be a digital bulletin board filled with copied press releases. Done right, it becomes a long-term SEO asset that consistently attracts readers month after month—provided you stop creating disposable URLs for recurring events.

A well-structured events hub, supported by smart internal linking, can generate reliable organic traffic regardless of the season. More importantly, it matches how people actually consume content today. Readers aren't settling in with a cup of coffee to read every word of a 1,500-word article—they're scanning for answers.

If your website can't tell someone what's happening this Friday night within a few seconds, Google will likely send that reader to another local publisher who can.

The good news? Building an AI-friendly, search-optimized events hub doesn't require a massive editorial overhaul. With a few structural changes and the right publishing workflow—and a little storytelling instinct—your newsroom can turn routine event listings into a dependable source of evergreen search traffic and content readers actually want to share.

How to Build an AI-Friendly Events Calendar

Traditional search results are no longer the only gateway to your content. As Google's AI-powered experiences continue to evolve, users increasingly receive complete answers before they ever click a blue link.

If someone searches for "things to do in Austin this weekend," Google's AI may summarize the best options directly in search results. The publishers that earn those mentions are usually the ones that present clear, structured information from the very beginning.

One of the biggest mistakes newsrooms make is burying the essential details halfway through an article. If readers—and AI systems—have to hunt for the event date, location, admission price, or start time, you've already made your content harder to understand.

Instead, follow the classic inverted pyramid.

Immediately below the headline, include a concise information box covering the essentials:

• What is the event?
• Where is it taking place?
• When does it start?
• How much does it cost?
• Who is it for?

This format doesn't just improve the reader experience. It also makes your content significantly easier for search engines and AI systems to interpret, increasing the likelihood that your information will appear in AI-generated answers, rich results, and other enhanced search features.

Long introductions, marketing language, and unnecessary storytelling may read well, but they often delay the information users are actually searching for. In an AI-first search landscape, clarity consistently outperforms cleverness.

Every event page should answer the user's primary question within the first few seconds. When your newsroom consistently publishes content that directly satisfies search intent, you build topical authority—a signal that has become increasingly important as search evolves beyond traditional keyword rankings.

This doesn't mean your events hub has to read like a spreadsheet. The facts come first—but the color and connection that make a story feel local can live comfortably right after the info box. See the next section for how to do both at once.

Give Readers the Facts—Then Give Them the Feeling

Structured data gets your event in front of readers. Good storytelling is what makes them stay, and what makes them trust your newsroom enough to come back for the next one.

The inverted pyramid isn't a reason to strip your event coverage of personality—it's a sequencing rule. Facts first, feeling second. Once the info box has answered what, where, when, how much, and who, the rest of the article has room to do what a press release never can: make the event feel like part of the community's life, not just an entry on a calendar.

Where cultural detail belongs, and where it doesn't

• Use it in the narrative body, never in the info box, headline, or structured data. Schema markup and the essentials block stay clean and literal—that's what search engines and AI summaries parse.
• Use it to answer a question the facts alone can't: why does this event matter here, and why now?
• Keep it brief. A sentence or two of texture is often enough. The goal is a reader feeling recognized, not a travel brochure.

A few examples of the kind of detail that earns its place:

• Independence Day: backyard barbecues, neighborhood fireworks, baseball games, small-town parades, concerts in the park, flags on front porches, families packing coolers for the lake—the unofficial midpoint of summer.
• Thanksgiving: crowded airports, family reunions, football rivalries, the start of Black Friday anticipation.
• Memorial Day: the unofficial start of summer travel season and ceremonies honoring veterans.
• Labor Day: the last beach weekend of the season and the shift into back-to-school routines.
• Election Day: campaign signs on lawns, polling places set up in school gyms and church halls, "I Voted" stickers.
• Christmas: tree lightings downtown, community toy drives, packed shopping districts.

Add context, briefly, when it helps a reader connect:

• A short line on why a tradition exists (a county fair rooted in the region's agricultural history, a parade that's run every year since a specific founding event).
• A nod to how the tradition has been passed down—"families have brought lawn chairs to this same stretch of Main Street for three generations."
• A connection between today's specific event and the wider ritual it belongs to.

Keep it grounded, not decorative. The test for any cultural detail: does it help a reader picture themselves at this event, or is it filler that could be swapped into any city's coverage of any event? If a sentence would read the same whether the event were in Ohio or Oregon, it's not doing its job. Specificity—this venue, this neighborhood, this year's twist on an old tradition—is what separates a locally reported guide from a generic listing, and it's also what keeps the writing from feeling like a press release run through a rewriter.

This is where a well-designed AI editing workflow earns its keep: it can be prompted not just to reorganize facts, but to notice where a sentence of local color belongs—and to leave it out everywhere else.

Keep Event Pages Working Long After the Event Ends

Here's how event coverage typically works in most newsrooms.

An editor publishes a preview of this weekend's event. The festival ends. The article stops receiving traffic. A year later, the newsroom creates an entirely new page for the next edition—and starts from scratch.

It's an incredibly common workflow, but it's also one of the biggest missed SEO opportunities for local publishers.

Every new URL created for the same recurring event dilutes your site's authority. Instead of strengthening a page that already has backlinks, search history, and rankings, you're asking Google to evaluate an entirely new piece of content every single year.

Publishing dozens of one-off event pages also wastes crawl budget and increases the risk of keyword cannibalization, especially when multiple pages compete for nearly identical searches.

Imagine your city hosts an annual county fair, a summer music festival, or a popular holiday parade.

Avoid creating URLs like:

yournewsroom.com/events/2026/summer-music-festival-2026

Instead, create a permanent, evergreen URL:

yournewsroom.com/events/summer-music-festival

The page becomes the long-term home for that event.

Before each new edition, simply update the date, schedule, featured performers, ticket information, and any other relevant details. Once the event ends, don't archive the page or let it go stale. Refresh it with a photo gallery, attendance highlights, key moments, or a recap of what happened.

That single URL keeps accumulating authority year after year.

Rather than abandoning everything you've already built, you're continuously improving an asset that search engines already trust. Readers searching for next year's event are far more likely to discover a page with an established history than a brand-new URL with no authority.

Think of every recurring event page as an evergreen resource—not a temporary news story.

An evergreen page also has a storytelling advantage the disposable model never gets: it can grow richer every year. This year's recap paragraph—the crowd size, a photo gallery, a specific moment from the parade route—becomes next year's "last year's festival drew a record crowd" line, building a sense of continuity a brand-new URL can never have on day one.

What If Local Organizations Don't Make Event Data Easy to Access?

Maintaining a comprehensive community events calendar by hand is one of the most repetitive jobs in any local newsroom.

Every day, editors copy information from city websites, museums, libraries, schools, theaters, nonprofit organizations, and community centers. It's important work—but it's also time that could be spent on original reporting.

Copying information from flyers, Facebook posts, or outdated websites isn't the best use of a journalist's expertise.

Fortunately, much of that workflow can be automated.

With Content Hub in 4Media CMS, publishers can aggregate external content feeds directly into their editorial workflow. New event announcements can be collected automatically, allowing editors to focus on reviewing, enriching, and publishing content instead of manually retyping it.

But what happens when a local theater or community center doesn't provide an RSS feed?

That's a surprisingly common problem.

Many smaller organizations still rely on basic websites—or simply post updates on Facebook or Instagram.

Fortunately, there are tools that can bridge that gap.

Services such as RSS.app, FetchRSS, and PolitePol can generate RSS feeds from almost any publicly accessible webpage or social media profile. Instead of checking dozens of websites every morning, your newsroom can automatically collect new event announcements in one place.

Once those generated feeds are connected to Content Hub, new content flows directly into your CMS, where it can be reviewed, enhanced with AI, and published in just a few clicks.

The result is a workflow that scales effortlessly—even if your community includes hundreds of organizations publishing events across dozens of different websites.

Give AI Better Instructions, Get Better Articles

Automatically importing event announcements into your CMS can save hours of manual work every week. But automation comes with one important responsibility: don't publish imported content exactly as you receive it.

If your CMS pulls an event description directly from a city website, library, museum, or community organization and republishes it word for word, you're simply duplicating content that's already available elsewhere on the web.

That creates little value for readers—and even less for search engines.

The solution isn't to abandon automation. It's to combine automation with AI-assisted editing.

Instead of treating AI as a content generator, think of it as your newsroom's first editor. Its job is to reorganize information, improve clarity, and transform raw event announcements into useful local guides.

A simple prompt can accomplish most of that work in seconds.

Example prompt

Act as an experienced local news editor. Transform the provided source material into an original, publication-ready article for a local news website. The source may be an event listing, press release, RSS item, webpage, social media post, or any other informational content. Preserve all verified facts while completely rewriting the text in original language and adding clear editorial structure.

An expanded version of the prompt, once your newsroom is ready to bring in cultural texture:

Act as an experienced local news editor writing for a community that knows this town well. Transform the provided source material into an original, publication-ready article. Lead with a concise facts block (what, where, when, cost, audience) before any narrative. Preserve all verified facts and never invent details, quotes, or attendance figures. After the facts block, you may add one or two sentences of authentic local or seasonal color—a relevant tradition, a seasonal routine, or a piece of context that helps a local reader picture the event—only if it fits naturally and stays factual. Do not add cultural references to the headline, facts block, or any structured data. If no genuine local angle is available, do not force one.

This approach does far more than rewrite sentences.

A well-designed AI workflow surfaces the information readers care about first, organizes it into a logical structure, and produces content that's easier for both people and search engines to understand—while still sounding like it came from a newsroom that knows the difference between a county fair in July and a harvest festival in October.

Instead of publishing another generic press release, your newsroom delivers a genuinely useful event guide while maintaining original, search-friendly content.

Help Search Engines Understand Your Events

Even the best-written event article is just plain text unless search engines understand what it represents.

To Google, an article about a summer concert, a county fair, or a charity fundraiser looks no different from a recipe or a restaurant review unless the page clearly identifies it as an event.

That's where structured data makes the difference.

By adding Schema.org Event markup behind the scenes, publishers give search engines the context they need to understand exactly what the page contains. This improves eligibility for enhanced search features, including rich results and AI-generated summaries.

With Events Widget in 4Media CMS, this process is built directly into the publishing workflow. Editors simply complete an event form, while the CMS automatically generates the structured data required by Google—no custom coding or developer involvement required.

One small detail is worth remembering.

Avoid adding marketing slogans or promotional copy to event titles and URLs. This applies to seasonal flourishes too—save the color for the body text, never the title, the URL slug, or the structured data fields.

Search engines prefer clarity over creativity. A title like:

Downtown Summer Music Festival

is far more useful than:

Don't Miss the Biggest Music Event of the Summer!

The same principle applies to event locations.

Always provide a real venue and physical address whenever possible. Accurate event names, locations, dates, and times don't just improve the reader experience—they also give Google's systems the structured information they need to confidently identify and display your event in search results.

In other words, you're not writing for robots.

You're giving search engines the context they need to understand what your newsroom has already created—and giving your readers the story around it.

Keep Your Events Hub Crawl-Friendly

Building a comprehensive events database is one of the best long-term SEO investments a local publisher can make.

But without proper technical maintenance, that same database can create unnecessary problems for search engines.

Google allocates a limited crawl budget to every website. The more time its crawlers spend exploring low-value pages, the less time they have to discover your newest stories and most important content.

This often happens when event directories include dozens of filter combinations, outdated archives, or dynamically generated search pages.

Imagine an events calendar where visitors can browse concerts from five years ago, combine multiple filters, sort by dozens of tags, and generate thousands of unique URLs.

To users, those filters may seem helpful.

To search engines, they're often an endless maze.

Instead of crawling your latest breaking news or upcoming event coverage, Googlebot may spend valuable crawl resources indexing thousands of low-value pages that have little chance of attracting organic traffic.

Fortunately, preventing this is straightforward.

Use your robots.txt file to block crawlers from unnecessary search results and dynamically generated filter pages.

Rules such as:

Disallow: /events/search

Disallow: /*?date=

Disallow: /*?tag=

help keep crawlers focused on the pages that matter most.

A clean crawling strategy allows search engines to spend their resources discovering fresh content instead of wandering through outdated archives and duplicate filter combinations.

The result is faster indexing, better crawl efficiency, and a healthier site overall.

Key Takeaways

A well-managed events hub isn't just another section of your website—it's a long-term traffic engine, and with the right editorial touch, a long-term relationship with your readers.

When built with both readers and search engines in mind, recurring event coverage can deliver consistent organic traffic throughout the year while strengthening your publication's authority in local search.

Here are the strategies every local newsroom should follow:

• Lead with the facts. Use the inverted pyramid and place the most important information—date, time, location, admission details, and audience—at the very beginning of every event page.
• Create original editorial value. Never republish municipal announcements or press releases verbatim. Use AI to transform source material into reader-friendly guides that provide unique value.
• Add authentic local color, sparingly. After the facts, a sentence or two of genuine seasonal or community context—why this tradition exists, how the town has marked it for years—helps readers feel connected without slowing down the search-and-AI-friendly structure.
• Use structured data. Help search engines understand your content by implementing Schema.org Event markup. Properly structured pages are more likely to qualify for rich search features and AI-generated answers.
• Build evergreen URLs. Instead of publishing a new page every year for recurring events, maintain one permanent URL and update it as each edition approaches. Over time, that page accumulates authority—and a running story of its own.
• Protect your crawl budget. Prevent search engines from wasting time on outdated archives and endless filter combinations. Keep crawlers focused on the pages that generate meaningful traffic.

Those are familiar shapes of decline. The newer and more useful part of this year's index is not how many reporters vanished. It is what fills the space they leave.

The Order of Collapse

A newsroom does not lose all its coverage at once. It loses it in a predictable sequence. As staff shrinks, editors keep the beats that reliably draw traffic and quietly drop the ones that take patient, time-consuming reporting. The civic beats go first.

The index measured this directly. Across 4.2 million articles published in the first three months of 2026, nearly 77% of U.S. counties produced zero local education stories that named a specific community. About 76% produced none on local health. These are not fringe topics. School closures, enrollment fights, hospital cuts, and insurance gaps are the decisions that shape a family's year, and in three out of four counties no reporter wrote a single local word about them.

Then comes the part that should unsettle every publisher. The vacuum does not stay empty. In counties with fewer than five journalists per 100,000 residents, nearly one in five local articles is about crime, against roughly 13% where reporters are more plentiful. That is about 50% more crime coverage, and not because there is more crime. It is because fear is the cheapest form of attention. When a shrinking newsroom cannot afford the slow work of covering a school board, it covers the police scanner instead. "If it bleeds, it leads" was always a tendency. Scarcity turns it into a business model.

What the Silence Costs

The consequences reach past the newsroom and into the ledger of the community itself. The index pairs its capacity data with civic outcomes, and while the authors are careful to call some of these correlations illustrative rather than proven, the pattern is consistent.

Local governments in states with below-average reporting capacity pay about 17% more in extra municipal borrowing costs, part of an estimated $1.1 billion that news deserts add to public borrowing every year. With no reporter watching the books, lenders price in the risk of waste, and taxpayers cover it. Residents of those communities also report measurably higher loneliness, three to eight points above comparably rural states with more journalists. And on a composite civic-participation score covering voting, volunteering, and giving, the most under-covered counties average 40 out of 100, against 58 for the best-covered. In the worst cases, the "double deserts" with almost no reporters and no outlets at all, that score falls to 23.

Strip out the local signal and a community does not just go uninformed. It goes numb: worse management, weaker oversight, lonelier residents, higher costs. One honest caveat belongs here. The study's content filter is strict and misses television, radio, and reporting that lives only on Facebook, so a few rural counties are covered slightly better than the raw counts suggest. The capacity trend itself, though, is not seriously in dispute.

What Reverses It

Here is the part worth holding onto: the decline is not uniform, and where it has turned around, the mechanism is clear.

Maryland now sits above the national average at 10.6 journalists per 100,000, largely because of The Baltimore Banner, a newsroom launched in 2022 that has grown to about 100 journalists funded by 82,000 paying subscribers. Baltimore now devotes roughly 8% of its local articles to health, three times the national rate. In Chicago, investment in outlets like Block Club has pushed Cook County's education coverage above the national average. These are not miracles. They are business models. A reader-funded newsroom can afford to keep the civic beats that an ad-dependent one is forced to drop.

That is the lever, and it is where we spend our time with publishers at 4media. The index makes the unit economics impossible to ignore: the beats that survive are the beats that pay, and right now crime pays in clicks while education does not. Changing that math means changing how a newsroom earns. A platform that lets a small team publish efficiently, convert readers into paying members, run newsletters and local events, and prove its audience to funders is what turns civic coverage from charity into something that sustains itself. We have made the case for revenue that does not depend on chasing the algorithm in Why Publishers Are Becoming Service Hubs and The $129 Million Scaffolding.

Technology does not conjure reporters. It decides which reporters a community can afford to keep.

Read Your Own County

The 2026 index lets you drop down to your own county and see exactly where your community stands. For a lot of American publishers, that map is uncomfortable reading. It is also the clearest brief you will get on where the need, and the opening, actually sit.

The silence is not permanent. Baltimore and Chicago are proof that capacity returns when someone builds a model that pays for it. Whether the desert keeps spreading is no longer the open question, the data has settled that. The open question is whether your newsroom becomes one of the places that proves it can be pushed back.

(Source: Local Journalist Index 2026, Muck Rack and Rebuild Local News)

The case began when two Munich publishers found that Google's AI Overview, triggered by searches for their names next to words like "scam," was telling users they ran subscription traps and dubious businesses and were tied to shady firms. None of it was true, and none of it appeared in the sources the AI cited. The system had confused them with genuinely sketchy companies, invented the connection, and served it as a confident answer at the top of the page. The court issued a preliminary injunction (case 26 O 869/26) barring Google from repeating the claims, under threat of fines up to €250,000.

The reasoning is the real precedent. A search engine that links to third-party pages is an intermediary, shielded by the liability protections platforms have leaned on for decades, including under the EU's Digital Services Act. A system that reads multiple sources and writes "entirely new, independent statements" in its own words is, the court found, producing its own content. Google argued that users can click through and check the sources for themselves. The court was unmoved: an AI Overview is a self-contained statement, barely 1% of users ever click a source link, and only Google can compare its output against the pages it drew from. The author answers for what it publishes.

What the Court Did and Did Not Say

It is worth being precise about the limits here, because the headline runs ahead of the ruling. This is a preliminary injunction from a regional court, not a final judgment and not a decision from Germany's highest court. Google has said it will appeal, and a full trial is expected later this year. The court stopped short of formally branding Google a "publisher" in the press-law sense; it ruled more narrowly that the AI Overview is Google's own statement, for which Google is directly liable. It deliberately broke from existing German case law on search engines, and a Frankfurt court reached a more cautious conclusion on a similar question last year. None of it binds a single American court.

What it does is puncture, in a courtroom and for the first time, the idea that an AI summary is just a neutral mirror of the web. As one media lawyer put the principle: if a platform's AI creates a statement, the platform is the author of it.

The Two Fronts Every Publisher Is Now Fighting

Step back and the position local publishers occupy looks strange. On one front, the whole industry is racing to get into these systems: AI visibility, LLM citations, "generative engine optimization," an entire advisory market grown up overnight, and the honest truth is that nobody can yet verify what reliably works. Everyone is measuring it as they go.

On the other front, those same systems can manufacture a damaging claim about you out of nothing. For a national brand, a hallucinated accusation is a problem. For a local outlet whose entire asset is the community's trust, an AI Overview that calls you a scam is closer to an existential threat. Your name is your business, and most local publishers have no communications team to fight a machine that defamed them above every search result.

The exposure is not hypothetical. An analysis for the New York Times found Google's current AI model answers correctly about 91% of the time, which sounds reassuring until you account for scale: at Google's search volume, the rest still means millions of wrong answers an hour. The same analysis found that 56% of even the correct answers could not be traced back to the sources cited. The machine is making claims its own citations do not support. The Munich ruling is the first sign that, in that fight, the law can land on the publisher's side.

What You Can Actually Do

You cannot file in Munich. But the ruling points at the real issue, control over how your name appears when a machine describes you, and there you are not powerless.

This is where we work with publishers at 4media, and three things matter. First, your own site has to be the authoritative, machine-readable source of truth about your organization: clean structured data, a clear masthead and "about" record, real author identities. When the ground truth about you is unambiguous and well-formed, an AI has less room to invent. Second, monitor what these systems actually say about your brand, so you can document harm the moment it appears, exactly as those publishers did before they went to court. Third, own your audience. The more your relationship with readers runs through channels you control, newsletters, memberships, direct visits, the less hostage you are to how any algorithm chooses to summarize you. We made that case in Google Search Live and the End of Drive-By Traffic.

A modern platform will not make you immune to a hallucination. It makes you legible to the machines and independent of them at the same time, which is close to the only durable position left.

Google's Move, and Yours

Google's next step is the question everyone is watching: tie AI Overviews tightly to the linked sources, add real correction tools, or fight the principle all the way up. Any of those would reshape how publishers appear online, and notably, binding the AI back to its sources is exactly what the publishers losing reach to AI Overviews have wanted all along.

You do not get to decide that. You do get to decide whether your outlet is easy for a machine to describe accurately and hard for one to misrepresent, and whether your business still stands if Google's summary of you is wrong tomorrow. Build for both. The era of "we just summarize, don't blame us" is ending. The work of controlling your own record is not.

(Source: Landmark German ruling on Google's AI Overviews (LG München I, case 26 O 869/26), The Decoder)

Modern language models process vast amounts of data in a matter of seconds, and then select only those sources they deem most credible and useful. It is precisely based on these that they create answers for users. If your website is not recognized as a valuable source of information, its content may remain invisible to potential clients.

Fortunately, this is not a process based on chance.

Visibility in AI-generated answers can be analyzed, measured, and systematically increased through properly planned optimization.

From SEO to GEO. What is the Paradigm Shift?

For years, the effectiveness of SEO activities was evaluated primarily using the Share of Voice metric. It determined what share of search engine traffic a given domain captured for selected keywords in traditional Google results, often analyzed alongside popular domain authority metrics like Ahrefs DR and Majestic TF.

Today, however, another metric is gaining importance - Share of Source, i.e., the share in sources utilized by artificial intelligence systems.

Share of Source shows how often content from a given website is indicated as a source of information and cited in AI-generated answers. While the goal previously was to achieve the highest possible position on the search results list, it is now crucial to provide information credible and valuable enough for the language model to use it when generating an answer.

With the development of AI-based search engines, the rules of optimization are also changing. Traditional activities, such as excessive keyword stuffing or expanding the backlink profile, are losing their significance for digital marketers figuring out how to survive and rank in the era of generative search.

GEO (Generative Engine Optimization) is playing an increasingly important role, which is an approach focused on preparing content and data structure in a way that makes it easier for AI systems to quickly find, interpret, and use them as a credible source of information.

Market analyses indicate that traffic coming from AI-based platforms, such as ChatGPT or Gemini, often features a higher conversion rate than traditional organic traffic from internet search engines.

This is primarily due to user intent. People using AI tools are usually looking for specific answers, recommendations, or solutions, rather than a list of pages to analyze themselves.

When an artificial intelligence system recommends a specific product, service, or company, and simultaneously bases its answer on information from a given website, the probability of performing the desired action - purchase, contact, or sending an inquiry - significantly increases.

Why Does the AI System Choose This Particular Article?

Most modern AI assistants, including ChatGPT, Gemini, or Perplexity, use solutions based on the RAG (Retrieval-Augmented Generation) architecture.

In practice, this means that the model first searches for information in available sources, and only then generates an answer for the user. However, before data is used, it goes through a quality and usefulness evaluation process.

One of the most important criteria is semantic relevance. The content should directly answer the user's question and provide specific information, without unnecessary digressions.

A clear site structure and architecture that Google loves is equally important for AI parsers.

Headings, tables, lists, and correctly applied HTML tags make it easier for AI systems to interpret and organize data. The timeliness of information and the presence of specific data, examples, and facts confirming the credibility of the content also matter.

For this reason, the BLUF (Bottom Line Up Front) method is playing an increasingly important role, which consists of presenting the most important information right at the beginning of a section or paragraph. AI models look for quick and unambiguous answers during content analysis, which is why it's worth placing key conclusions, definitions, or recommendations right at the beginning. The elaboration, additional explanations, and broader context should only be included later in the text.

The robots.txt File - How to Manage AI Bot Access?

One of the first steps in optimizing a website for AI search engines - alongside ensuring Google can discover your portal via a modern sitemap - is to check if the robots responsible for indexing and searching content have access to the site (i.e., what is the content of the robots.txt file located on the server).

Even valuable and well-optimized materials will not appear in AI-generated answers if the appropriate bots are blocked at the server level. This is a technical hurdle similar to diagnosing why a site is blocked in certain regions even without illegal content. The basic access control tool is the robots.txt file located in the site's root directory.

Many misunderstandings have arisen around AI bots.

Some website owners block all robots associated with artificial intelligence, wanting to prevent their content from being used to train models. However, this approach can unknowingly limit the site's visibility in AI-generated answers. It's crucial to understand that not all bots serve the same function.

For example, GPTBot is primarily used to acquire data utilized in developing and training OpenAI models. On the other hand, OAI-SearchBot is responsible for searching and retrieving content used when presenting results and answers to users.

PerplexityBot plays a similar role in the Perplexity ecosystem. Blocking robots responsible for searching can cause content to no longer be included in answers generated by AI tools, even if it remains visible in traditional search results.

To check the configuration, log into the hosting panel or connect to the server using an FTP client. Then, locate the robots.txt file located in the domain's root directory, usually in the public_html folder. It's worth making sure it doesn't contain rules blocking access to robots responsible for searching and indexing content.

If you care about restricting the use of materials for training AI models, but at the same time want to maintain visibility in AI-generated answers, you can block training bots while leaving access for search bots. A properly configured file should include the following directives:

User-agent: OAI-SearchBot Allow: /

User-agent: PerplexityBot Allow: /

User-agent: GPTBot Disallow: /

After saving the changes, it's worth checking the correctness of the configuration and regularly monitoring the documentation of AI service providers.

The development of AI-based search engines is very dynamic, so the list of used bots and their functions may change over time.

The most important rule is simple: blocking bots responsible for training models does not have to mean giving up visibility in AI answers.

To retain the chance of appearing in results generated by artificial intelligence systems, you must ensure that robots searching and indexing content still have access to the website.

The llms.txt File - An Additional Content Map for AI Systems

Traditional website HTML code contains many elements that have no meaning for language models. Navigation menus, analytics scripts, ads, CSS styles, or UI components are essential for users but make it difficult to quickly reach the actual content, especially if your CMS choice slows down the site and kills your SEO.

For this reason, the concept of the llms.txt file, which acts as a simplified guide to the site's most important resources, is generating increasing interest.

The llms.txt file is placed in the site's root directory, just like robots.txt. Its task is to point to the most valuable sections of the website in an organized, easy-to-process format.

Instead of analyzing the entire site structure, the AI system can more quickly identify pages containing key information, documentation, guides, or expert articles.

In practice, the file contains a list of the most important URLs - which should ideally follow a logical and optimized structure - along with a short description of their content. The simple Markdown format is most often used for this, which is readable by both humans and text-processing systems.

How to Create an llms.txt File?

Preparing the file is relatively simple and does not require advanced technical knowledge. First, create a new text file and save it under the name llms.txt.

Then, add links to the most important sections of the site along with a short description of their content.

The sample content of an llms.txt file might look like this:

XYZ Company Website

Knowledge Base

Knowledge Base Expert articles on SEO, GEO, and internet marketing.

Documentation

Documentation Instructions, guides, and technical materials for clients.

After preparing the file, it must be placed in the site's root directory so that it is accessible at your main domain address followed by /llms.txt.

Although llms.txt does not replace an XML sitemap or standard SEO activities, it can serve as an additional signal making it easier for AI systems to find the most important content.

In the case of extensive websites with a large number of articles, documentation, or expert materials, it also allows for better organization of resources intended for indexing and analysis.

Re-architecting the Technical Foundation: The Imperative for CMS 4Media

The transition from SEO to GEO necessitates a rigorous reevaluation of a publisher's underlying technological stack. Legacy Content Management Systems (CMS), particularly open-source platforms bloated with decades of plugins, unstructured code, custom widgets, and heavy client-side scripts, actively impede AI retrieval. LLM crawlers operate with strict context windows and parsing limits. If a crawler is forced to navigate a labyrinth of inline CSS and complex JavaScript to locate the core content, it will likely abandon the crawl entirely.

To capitalize on GEO, publishers require an agile, structurally clean, and media-rich platform. This is where specialized infrastructure like CMS 4Media becomes a distinct competitive advantage.

Developed as a comprehensive, enterprise-grade ecosystem tailored exclusively for digital publishers, local media, and broadcasters, CMS 4Media natively addresses the mechanical requirements of Generative Engine Optimization through several core architectural advantages:

• Structured Content and Semantic Clarity: CMS 4Media employs an advanced visual widget system that enforces a clean, modular hierarchy. Editors can construct complex layouts without touching the underlying code, ensuring the HTML output remains pristine and highly readable for AI parsers.
• GEO-Optimized Workflows: The text editor is designed specifically for modern publishing. It includes dedicated fields for content "Introductions" that perfectly align with the BLUF (Bottom Line Up Front) methodology. These leads are automatically formatted and wrapped in specific HTML markers that explicitly communicate with search robots, identifying the most vital, extractable summaries required for RAG systems.
• Rich Multimedia Contextualization: As generative models become increasingly multimodal, contextualizing video and audio is paramount. CMS 4Media features a robust multimedia module that creates dedicated subpages for uploaded assets, allowing editors to append detailed titles, descriptive metadata, and WebVTT subtitles. This provides AI crawlers with deep, text-based context, drastically increasing the likelihood that a publisher's rich media will be referenced in AI answers.

For media organizations looking to survive the transition to generative search, upgrading to an AI-ready infrastructure like CMS 4Media ensures that high-quality journalism is perfectly formatted for the algorithms that now distribute it.

How to Measure AI Search Engine Traffic in Google Analytics 4?

One of the biggest challenges associated with optimization for AI search engines is measuring the effects of the conducted activities. Proper tracking requires a solid understanding of web analytics, similar to setting up your workspace with a technical guide to Google Analytics 4 and Search Console.

Unlike traditional search engines, traffic coming from tools such as ChatGPT, Perplexity, or Gemini is not always unambiguously classified in analytics reports.

Some visits pass on source information, while others may be assigned to Direct, Referral, or Unassigned channels, which makes it difficult to assess the real impact of traffic generated by AI.

To get a more complete picture of the situation, it's worth creating a dedicated channel group in Google Analytics 4 for visits coming from the most popular AI-based platforms. This allows you to monitor the number of sessions, user behavior, and conversion rates for this traffic source.

How to create an "AI Traffic" channel in GA4?

In the Google Analytics 4 panel, go to the Admin section, and then select Channel groups under the Data display area.

You can create a new channel group or edit an existing custom group. Click the "Create new channel group" button.

Next, add a new channel, giving it a name, for example, "AI Traffic".

In the conditions configuration, select the Session source parameter, and set the operator to matches regex.

Depending on your needs, you can create one collective channel covering all traffic from AI platforms or separate channels for each tool.

The first solution allows you to quickly assess the scale of AI-generated traffic, while the second makes it easier to analyze the effectiveness of individual platforms.

For the collective "AI Traffic" channel, you can use a regular expression such as:.*chatgpt.*|.*openai.*|.*perplexity.*|.*claude\.ai.*|.*gemini.*|.*copilot.*

If you care about more detailed data, create separate channels for individual traffic sources using partial expressions:

• ChatGPT/OpenAI: .*chatgpt.*|.*openai.*
• Perplexity: .*perplexity.*
• Claude: .*claude\.ai.*
• Gemini: .*gemini.*
• Microsoft Copilot: .*copilot.*

During channel configuration, select the "matches regex" condition, and then paste the appropriate pattern.

Special attention should be paid to the order of rules in the channel group.

Google Analytics analyzes them from top to bottom and assigns the session to the first matching channel.

After creating the channel, click the "Reorder" button located next to the channel list. Move the newly created channel, and confirm the decision with the "Apply" button.

This means that the newly created "AI Traffic" channel should be above general rules, such as Referral.

Otherwise, some visits will be qualified into other categories, which will hinder later analysis.

After saving the configuration, data will start appearing in reports according to the new classification rules. Thanks to this, it will be possible to monitor what share of traffic and conversions is generated by users landing on the site via artificial intelligence tools.

Where exactly to look for this information?

To find the new group in the Google Analytics 4 panel, select Reports from the menu on the left, then expand the Acquisition (Lead generation) section and click Traffic acquisition. In the main data table, you will find your newly created "AI Traffic" channel.

Tools for Monitoring Citations. From SaaS to Python

Occasionally checking brand visibility by manually entering queries into ChatGPT, Gemini, or Perplexity only allows you to get a general picture of the situation.

However, this approach is hard to consider a reliable measurement method.

Answers generated by language models can vary depending on the context of the conversation, current data sources, the phrasing of the question, or the search mechanisms used.

For this reason, effective monitoring in the GEO area requires a more systematic approach.

More and more tools specialized in analyzing brand visibility in AI-generated answers are appearing on the market.

Platforms such as Otterly.ai or Profound automatically send sets of predefined queries to various language models, and then analyze the answers for the presence of indicated brands, products, or domains.

A free, budget-friendly alternative to start with is to use a free account in the Ahrefs tool (Ahrefs Webmaster Tools). You just need to add your domain there as a project.

In the basic reports, a list of AI systems in which citations of your website appeared will be visible.

However, it should be remembered that this is a very basic version - to see more detailed data, such as specific subpages used as a source or exact citation dates, it is necessary to upgrade to a paid package.

Thanks to this, it is possible to track changes in visibility over time and identify topics where the competition is more frequently cited as a source of information.

Companies with technical resources can also create their own monitoring system. In practice, this involves using APIs provided by AI model providers to automatically ask questions and analyze the obtained answers.

These types of solutions allow you to monitor specific keywords, check the frequency of brand occurrences, and identify sources indicated by models when generating answers.

The greatest advantage of automation is the ability to regularly collect data and observe trends.

Thanks to this, it's easier to detect areas where the brand is rarely cited, and then supplement the content with missing information or expand sections answering users' specific questions.

In practice, this allows you to react faster to changes occurring in AI search engines and more effectively increase the site's share in sources utilized by language models.

FAQ Section

How does AI optimization (GEO) differ from classic SEO?

SEO and GEO share a common goal - increasing content visibility on the internet - however, they use slightly different mechanisms.

Classic SEO focuses on improving the site's position in search results by optimizing content, site architecture, user experience, and the link profile. GEO (Generative Engine Optimization) focuses on increasing the probability that the content will be used as a source of information by artificial intelligence systems.

In practice, this means a greater emphasis on clear information structure, unambiguous answers, credible data, and formats that facilitate analysis by language models. It's no longer just about taking a high position in search results, but also about the content being recognized as a valuable source when generating an answer by AI.

Why does ChatGPT point to a competitor's site more often than mine?

There can be many reasons. Language models and AI search systems prefer content that clearly answers users' specific questions.

If competitors publish more organized materials, use clear headings, sections with key takeaways, tables, lists, and data backed by sources, their content may be easier to interpret and cite.

Domain authority, the timeliness of published information, brand presence in other credible sources, and the way knowledge is presented also matter.

In many cases, improving the overall content quality and applying technical on-page fixes - such as using the proper canonical tag in sponsored content to consolidate link equity - can significantly increase the chances of appearing in AI-generated answers.

Will blocking GPTBot in the robots.txt file cause the site to disappear from ChatGPT results?

No. GPTBot is primarily responsible for acquiring data used to develop and train OpenAI models.

Blocking it does not mean automatic removal of the site from search results or answers generated by ChatGPT.

If you want to restrict the use of content for training purposes while maintaining visibility in search, it is crucial to leave access for bots responsible for indexing and retrieving content for search purposes, such as OAI-SearchBot.

These are the ones that can use the website's content when creating answers and presenting sources to users.

Key Takeaways

• Share of Source is gaining importance as a complement to the traditional Share of Voice metric. It's becoming increasingly important not only to occupy high positions in search results, but also to be present in sources utilized by AI systems.
• The robots.txt file requires conscious configuration. Blocking bots responsible for training models does not have to affect website visibility, however, blocking searching and indexing robots can limit its presence in AI-generated answers.
• The llms.txt file can make it easier for language models to find the most important content on the site and is a valuable addition to GEO-related activities.
• A modern technical foundation is essential. Upgrading to an AI-ready infrastructure like CMS 4Media ensures clean HTML, GEO-optimized workflows, and proper media contextualization, bypassing the limitations of legacy platforms.
• Traffic coming from AI platforms is not always unambiguously visible in Google Analytics 4. It's worth creating custom channel groups to more effectively monitor this user segment.
• Optimization for AI search engines is based on a clear content structure, exposing the most important information at the beginning of a section, and providing specific, credible data.
• GEO does not replace SEO, but extends it with activities increasing the chance of using content as an information source in answers generated by artificial intelligence, which is equivalent to the classic positioning process in the AI era.

He is not a coastal fluke. Wichita has Wichita Life. Upstate New York has Catskill Crew. London has Wooden City, with more than 12,000 subscribers, enough to support its writer full-time. The New Yorker recently profiled the whole wave. The pattern holds across very different markets, and the reason is not better writing. It is sharper focus, a visible human, and relentless local utility.

This is not a story about Substack replacing local journalism. It cannot. Real newsrooms have reporters, archives, sources, and decades of community trust that no part-time writer can assemble. But these solo operators are doing five things, visibly and profitably, that a lot of American local publishers have stopped doing. They are worth studying.

Market Data Brief: The Solo Operator Advantage

• Conversion gap: The Eastside Rag converts roughly 20% of its 5,500-plus subscribers to paying, while most newsletters convert only five to ten percent.
• Engagement premium: Beehiiv reports neighborhood newsletters earn a 2.5x higher click-through rate than the platform average.
• Audience scale: Wooden City has passed 12,000 subscribers on a single-city beat and now pays its writer a full-time living.
• Revenue logic: Paid newsletter subscriptions are recurring, first-party revenue. They do not depend on ad networks, search algorithms, or third-party cookies, which are the exact revenue sources draining out of local media.

Five Things Solo Operators Get Right

1. A face, not a masthead. Every one of these newsletters is one named person. Readers know who walked past the new café, who sat through the council meeting, who holds the opinion. The institutional voice is losing to a single human with a byline.

2. Utility before everything. Isaac Rangaswami's rule for Wooden City is that every issue has to be practical enough for people to pay for it. Where to eat. Where to park for the game. Which contractor actually showed up on time. When Google answers a local question with an AI summary or a wall of SEO sludge, a person who has actually been there is the entire differentiator.

3. Narrow beats broad. The instruction is to go smaller than feels comfortable. Not the whole metro, but one side of the highway. Not the county, but the school district. That focus is what produces the engagement premium. A tight beat creates mail people actually open.

4. Opinion is an asset, not a liability. These writers take sides, on a bike lane, a zoning variance, a beloved diner that closed. Hard news reporting still needs to run straight. The voice around it does not. Neutrality reads as absence, and if no reader ever writes in angry, no reader is invested.

5. Design that signals place. Alexa Tietjen Dornagon built Court Street Journal to feel like a storybook, with watercolor illustrations of local storefronts. It looks like the place it covers. Most local news sites look interchangeable, and that sameness is a positioning failure. Aesthetic is positioning.

What Real Newsrooms Already Have

Here is the part the "we can't compete with Substack" argument misses. Established local publishers already hold the assets a one-person operation will never build: trained reporters, a searchable archive, real sources inside city hall, and years of accumulated trust. The solo operator is improvising the credibility your newsroom earned decades ago.

What you are missing is not capability. It is the willingness to write, design, and edit like one person who actually lives on the block.

The Infrastructure That Decides It

It's where we hit the same wall constantly when auditing publishers at 4media. A newsroom decides to put named writers up front, build a genuinely useful local beat, and give the site a look of its own, and the legacy CMS fights every step.

Author identity gets buried under a generic template. The newsletter tool cannot segment a neighborhood list or gate paid content without a third-party bolt-on. The design is locked to a stock theme shared by a thousand other sites. When publishing like a human requires a custom development project, the initiative stalls before the first issue ships.

A modern platform treats these as core features, not optional add-ons. CMS4media gives every writer a real author profile and byline, runs newsletters and paid membership natively, and lets editors shape the look of the site so it reflects the place it covers. Whether your audience ever turns into recurring, first-party revenue is, in the end, a product decision, and your software either enables it or blocks it.

Take It as an Opportunity, Not a Threat

The town crier is back, and the lesson is not that local publishers have been beaten. It is that the winning playbook is now obvious, visible, and cheap to copy.

You already have the reporters, the archive, the sources, and the trust. Put a face on the front page. Make every issue useful enough to pay for. Go narrower than feels safe. Hold an opinion. Make the site look like the place it serves. Then make sure your technology can actually deliver all five, because right now that is the only thing standing between your newsroom and the conversion rates a part-time writer is posting before breakfast.

We traced the collapse of search referral traffic in Google Search Live and the End of Drive-By Traffic and The End of "Back Button Hijacking". The through line is the same: the open web is no longer a reliable landlord. The audience you own is the only one you can count on.

(Source: Your Friendly Neighborhood Newsletter, The New Yorker)

FAQ

How many subscribers does a local newsletter need to be profitable? Far fewer than most publishers assume. When the beat is tight, engagement runs high: Beehiiv reports neighborhood newsletters get 2.5x the click-through rate of the platform average. Solo operators are sustaining themselves on audiences in the low thousands, because a focused local list converts to paid at much higher rates than a broad one.

What paid-conversion rate should a local newsletter expect? Most newsletters convert five to ten percent of free subscribers to paying. The strongest neighborhood newsletters, such as The Eastside Rag, reach roughly 20%. The difference comes from focus, a named human voice, and consistent utility, not from a larger audience.

Can an established newsroom compete with a solo Substack writer? Yes, and it starts ahead. A newsroom already owns reporters, archives, sources, and community trust a part-time writer cannot replicate. The gap is execution and tooling: a CMS that supports real author identity, native newsletters, paid membership, and a distinct design.

How to Activate These Features?

Go to: Settings module → Ad Block Detection

The first available option concerns the audio/video player.

Activating this feature means that a user who is actively using ad-blocking software will see a message informing them that video playback has been blocked when they click on the player. The restriction will be lifted once the user disables the ad blocker for the website they are browsing.

The second available feature is an informational banner, which is automatically displayed to users using ad-blocking software.

Clicking the Show information banner button expands additional settings.

There are two available modes:

• Show banner but do not block access to the portal - with this option enabled, the user will see a notification informing them about detected ad-blocking software, but they will still be able to continue browsing the website.

• Show banner and block access to the portal - with this option enabled, users with an active ad blocker will see a banner informing them that access to the portal is blocked. The restriction will be removed automatically once the user disables the ad blocker. After clicking Refresh page, the user will be able to access the portal without restrictions.

Additional Settings

When enabling the banner feature, two text fields must be filled in: banner title and banner content. It is the publisher’s decision what message will be displayed to users with an ad blocker.

Additionally, the portal administrator can configure the visual layer. Using available settings, it is possible to define the font color and background color of the banner. There is also an option to enable a border, where its color, thickness, and style can be defined.

In this layout, users browse photos by scrolling vertically. After scrolling through a set number of images, the browser tab refreshes (reloads) to display the next batch of photos in the gallery.

This solution allows for the seamless display of banner ads between photos, while ensuring page views are accurately tracked during the refresh triggered by scrolling. Ads are served using the Type C ad widget.

How to activate the Hybrid Gallery:

• Navigate to Settings -> Photo Galleries.
• In the Gallery Type field, select Hybrid Cascade Gallery.
• Define the number of photos after which the browser tab should reload and load the next set of images.
• Save your changes by clicking the Save button.

Hybrid Gallery Interface

After viewing the first group of photos, the user will see a "Next Page" button. They can either click it to reload the tab or simply scroll further down to trigger an automatic refresh, which will load the next batch of images.

Users can return to the previous set of photos by clicking the arrow button.

Below the last image in the gallery, the user can use the Previous page button

To activate ad widgets displaying C-type banners between images in the gallery:

• We log in to the Ads4Media advertising platform
• in the CMS4media Portals tab, click the Options button and then the Ad spots button
• click the green Add new C ad space button
• select the option Available on the lists, in between the elements
• from the expanded options, select  – List of photos in the cascade gallery
• specify after how many images the ad widget should appear for the first time and whether it should be displayed again – if so, specify every how many images its display should repeat

As part of the cascading gallery and hybrid gallery, a solution has been implemented that enables the activation of independent C widgets displaying central ads as well as separate widgets with local ad delivery.

To customize the watermark settings, go to the [Settings] > [Multimedia] module.

Further down the page, a new section named "Watermark" has been added, where you will find the following personalization options:

1. Activating the watermark feature per module and setting its default status
   
   Each module includes two dedicated toggles:
   1. Enable the watermark – determines whether the watermark feature is active in a given module.

   2. Enabled by default – an option available only when the first toggle is enabled. It allows you to decide whether the watermark should be automatically activated after adding a main image.

      If the [Enabled by default] option remains disabled, the watermark must be activated manually within a specific entry. This solution is useful when only selected publications use the watermark functionality.

2. Watermark – a field to select the watermark graphic. It must first be uploaded to the portal's file library; supported formats are png and webp.
3. Watermark - location– a grid interface that allows you to adjust the position of the watermark on the photos.
4. Vertical / Horizontal spacing / Maximum width – in addition to the position, you can adjust the margins from the edges and the size of the watermark. All values are expressed in pixels.

After configuring all settings, remember to save your changes at the bottom of the page by clicking the dark blue "Save" button.

Configuring the watermark for a featured image in a specific entry

In addition to the global settings described above, you can also customize the watermark position for the featured image of a specific post. This applies to all modules where this functionality is available.

We will walk through the process using the [Articles] module as an example. Other modules follow the exact same method.

In the article edit view, within the [Multimedia] section, after uploading a featured image, the following will appear:

1. "Enable the watermark" toggle along with a grid to adjust its position for this specific image. By toggling this switch, you can also completely deactivate the watermark for this graphic.
2. A preview of the uploaded main photo showing the watermark visualization.

If you already have a featured image uploaded and wish to update it, the preview of the new graphic will be displayed below the currently uploaded one.

Watermarks in additional photos and galleries

These cases automatically follow the global configurations set in [Settings] > [Multimedia].

IMPORTANT: Watermark settings do not work retroactively. This means that new settings, as well as a newly uploaded logo, will only be applied to new entries or after manually editing existing ones. The same applies to additional photos and individual images in galleries.