Preferred Sources: eligibility, effect, and what it counts

Google now takes a ranking input directly from your reader. The Preferred Sources button is code a publisher embeds on their own pages. When a reader clicks it, the site is added as a preferred source on that reader’s Google account, and their Top Stories, AI Overviews and AI Mode results start favoring it from then on.¹ Google says more than 600,000 unique sources have already been selected.¹ Before any of that matters to you, check whether your site is eligible, because Google’s documentation excludes an entire class of publisher that the announcement never mentions.

Only domain-level and subdomain-level sites are eligible. A subdirectory is not.

What is Google’s Preferred Sources button?

Publishers embed the Preferred Sources button as an interactive control on their own pages, so readers can mark them as a preferred source on Google. Google’s announcement describes the flow: a reader clicks, the site is added to their account, and they are returned to where they left off on the page.¹ Readers then find those publications more easily in Top Stories, AI Overviews and AI Mode.¹

The shape of that is new for anyone working on these surfaces. Schema, Core Web Vitals and AMP were all things you did to your own pages, and each one paid off through a system evaluating what you shipped. This one does nothing until a person acts. You can put the button on every page you own and end the quarter with zero selections, because the mechanism is a request rather than an implementation.

Three implementation paths exist, and they matter less than the eligibility question below. The standard JavaScript route is two lines producing an automatically localized, Google-styled button, which Google recommends. An advanced route lets you use your own design assets. A deeplink, which is a plain link that triggers the same action without any code on the page, covers sites whose content management system (CMS) cannot run the button, and it works in social posts and newsletters too.²

Which sites are eligible, and which are not?

Only domain-level and subdomain-level sites are eligible to appear in Google’s source preferences tool, and the documentation draws the boundary with its own worked example. A domain such as https://www.example.com/ qualifies, a subdomain such as https://code.example.com/ qualifies, and a subdirectory such as https://www.example.com/blog does not.²

That single sentence disqualifies a large group of publishers, and it appears nowhere in the announcement. A brand publication living at /blog, /insights or /newsroom on a corporate domain cannot be selected as a preferred source, no matter how much original reporting sits there. The eligible entity is the whole domain, which for most companies means the commercial site rather than the editorial operation inside it.

Google frames the check as a lookup instead of a rule to reason about. Enter your site in the source preferences tool’s search box, and if it appears, the implementation paths apply.² Do that before writing a ticket, because the alternative is discovering the exclusion after the work is scoped.

What does a preferred-source selection actually do?

A selection earns a badge and an unquantified lift. Google’s documentation says content from a selected site is “more likely to appear” in Top Stories, highlighted with a preferred badge, and that in AI Mode and AI Overviews it “can be highlighted” with the same badge for users who selected it.² Availability is broad: Top Stories in every language where Google Search runs, plus AI Mode and AI Overviews wherever those features are available.²

Read those verbs at their real strength. “More likely” and “can be highlighted” are hedges Google chose deliberately, and the announcement’s framing of publishers gaining a more seamless way to connect with readers sits some distance above them.¹ The effect is also scoped to the individual: a badge shown to a reader who already chose you is a smaller claim than a ranking change, since that reader had already demonstrated the preference the badge now serves.

In my opinion, the personalization framing is the honest one. Google is not promoting you to a wider audience, and it is making you easier to find again for people who went out of their way to say they wanted you.

What does the 600,000 figure count?

Google’s 600,000 figure counts unique sources selected at least once, and nothing else.¹ Google’s wording is precise: “So far, people have already selected more than 600,000 unique sources.”¹ That is a count of distinct publishers who have received at least one selection from at least one person.

Three things it does not tell you. It does not say how many people did the selecting, so one enthusiastic user and a million users produce the same figure for a given site. It does not say how many selections each source received. And it says nothing about impressions, clicks or traffic, which is the quantity a publisher would actually need to price the work.

I read 600,000 as a small number against the scale of Google’s index. Its usefulness is as evidence that the feature is live and being used, and it will get quoted as a reach number by people who did not read the sentence around it.

Does Preferred Sources help a site that is not a news publisher?

Google has not said whether Preferred Sources helps a site that never appears in Top Stories. The feature is documented against Top Stories first, with AI Mode and AI Overviews named as additional surfaces where the badge can appear.² Top Stories is a news surface, and neither source addresses what happens for a publisher who never appears there.

For most e-commerce and technical audiences that is the question that decides whether to bother, and the honest answer today is an absence, not a no. The AI Mode and AI Overviews language is not scoped to news in the documentation, which leaves room for a non-news site to benefit. Suspected, not documented, and I would not build a case on it.

There is a reasonable read available without resolving that. If your audience returns to you directly and you publish often enough for recency to matter, the button costs two lines and asks nothing of your architecture. If neither is true, the mechanism has no fuel, because a signal that only fires when a returning reader clicks is worth roughly what your returning-reader base is worth.

Is the Preferred Sources button worth adding?

Adding the Preferred Sources button is worth it on an eligible domain with a returning audience, as a cheap experiment and not a strategy. The work is small, the risk is nil, and the upside is real for exactly the readers who were already going to come back. Treat it the way you would treat any low-cost addition whose payoff Google has declined to quantify.

Three things to do in order. Check the source preferences tool for your domain first, because everything downstream depends on that answer. If you are eligible, add the standard button and put it where a returning reader will actually see it, which is the end of an article rather than a global header. Then leave it alone, since there is no reported metric to optimize against and no dashboard that will tell you what it did.

If your publication lives on a subdirectory, the decision is made for you. Spend the time on something the documentation says you can affect.

One caution about the next few months. This launched on August 20, 2026 and its documentation is new, so the eligibility rules are the most likely thing to move. Re-read the Search Central page before acting on anything here, because the strongest claim in this article is also the one Google could revise without announcing it.