How to Use Google's Preferred Sources Button to Boost Your AI Overview Rankings

What changed with Google's Preferred Sources button?

Google released an interactive Preferred Sources button that publishers can embed on their pages. A reader who clicks it can add the publisher's site as a preferred source in Google, then return to the page they were reading. The product change matters because the action no longer requires a reader to navigate through Google's source-preference settings on their own.

The August 20, 2026 release changed the path from a settings-led choice into an on-site interaction. Google describes the standard implementation as two HTML additions: a script that loads Google's library and a placement element where the localized button renders. Publishers that cannot add the interactive version can still use a source-preferences deeplink in a page, newsletter, or social post.

Google had already brought Preferred Sources into AI Overviews and AI Mode on May 27, 2026. That earlier update let users identify links from sites they had selected through a Preferred label. The newer button does not replace the AI Search feature. It makes it easier for a publisher to invite a reader to use it.

That distinction is essential for anyone working on AI search visibility. The release changes how a reader can express trust in a source. It does not establish a new technical method for forcing a URL into an AI Overview. Google's publisher documentation says selected content can be highlighted with a Preferred badge in AI Mode and AI Overviews for users who selected the site.

1. Before May 27, 2026: Preferred Sources was associated with a reader's source preferences and Top Stories.
2. May 27, 2026: Google said Preferred labels would appear for selected sources in AI Overviews and AI Mode.
3. August 20, 2026: Google released an embeddable button that lets readers add an eligible site and return to the originating page.

Does the Preferred Sources button boost AI Overview rankings?

The Preferred Sources button can improve how a selected source is presented to a reader, but Google does not say it raises a site through AI Overview rankings for everyone. The documented AI effect is a Preferred badge on content for users who have already chosen that site. Treat that as a personalization and click-through opportunity, not a promise of broader answer inclusion.

This is where the language around the update gets sloppy. A publisher may reasonably want more appearances in AI answers, yet a reader's preference and an engine's decision to cite a page are different signals. Citation Engineering focuses on the latter through quality, coverage, and freshness. The button can create a stronger path with an audience that already knows the publication, while the underlying work still needs to make each page worth citing.

Google does describe a stronger result in Top Stories: content from a preferred source is more likely to appear and is highlighted with a Preferred badge. The AI Overview and AI Mode wording is narrower. Google says the selected source's content can be highlighted with that badge for that reader. The difference is not a technicality. It is the boundary between an official product capability and a ranking claim that Google has not made.

Google reported on May 27 that people were twice as likely to click through to a Preferred Source. That is a useful reason to test the reader experience, especially for publishers with recurring audiences. It is not evidence that every click adds an AI Overview citation, that the effect applies to unselected readers, or that a button can compensate for weak source material.

The practical response is to separate two measurements. Track button adoption or source-preference engagement as an owned-audience signal. Track Citation Share separately across the relevant prompt set, because Citation Share measures the percentage of relevant AI answers in a category that cite you. A gain in one measure should not be presented as a gain in the other without direct evidence.

Who is affected by the new button?

Publishers and site owners with eligible domains are affected first, because Google makes the button available to websites that appear in its source-preferences tool. The has is global for Top Stories wherever Google Search is available. Preferred Sources can also appear in AI Mode and AI Overviews in the languages and locales where those AI has are available.

Eligibility has a hard structural boundary. Google says domain-level and subdomain-level sites can appear in the source-preferences tool. A root domain such as example.com and a subdomain such as news.example.com can qualify, while a path such as example.com/news cannot. This matters for businesses that publish content in a folder on a main marketing domain: they should test the actual domain in Google's tool before assigning development work.

The button is most relevant where a publisher has readers who choose to return. A specialist publication, a local newsroom, a research-led B2B blog, or a service business with a trusted resource center may have a natural reason to ask readers to add it. A site with no repeat audience should not treat the button as a shortcut. The has asks a person to make an explicit preference choice.

It also affects editorial and growth teams that report AI performance. A visible Preferred label can alter a reader's choice once a link appears in an AI response. That makes the has relevant to click behavior and brand trust. It does not remove the need to publish direct answers, maintain current pages, and provide sources that support the claims being made.

For Omnicite's clients, the useful framing is simple. The button may strengthen the connection between a known reader and an eligible domain. It sits beside work that aims to earn citations across ChatGPT, Perplexity, Gemini, Copilot, and AI Overviews. It is not a replacement for an editorial system designed around the questions buyers actually ask.

How should publishers respond to the Preferred Sources release?

Publishers should verify eligibility, install the lowest-friction version they can support, and then measure the result without overstating it. Begin at Google's source-preferences tool and search for the precise domain or subdomain that readers should select. If it does not appear, do not ship a call to action that cannot complete its intended action.

If the site is eligible and the CMS supports the required code, use Google's standard interactive implementation. Google recommends it because it provides a localized, Google-styled button and returns the reader to the page after selection. Place it where a reader has already received enough evidence to trust the publication, such as after the answer-first summary, near a newsletter invitation, or in an author and methodology area.

If the CMS cannot host the script, use Google's deeplink option. The fallback is materially less integrated, but Google explicitly supports using it as a text link or clickable image, including in social posts and email newsletters. The choice is operational rather than ideological: use the version that your reader can complete and that your team can maintain.

Keep the call to action plain. Explain that it helps readers find the publication more easily in Google and that it is optional. Do not say the reader will make the site rank first in AI Overviews. Do not tell them the action changes all Google results. Those claims exceed Google's public documentation and risk turning a trust has into a misleading growth tactic.

Finally, run a bounded experiment. Compare the number of readers who use the button or deeplink with downstream engagement on the pages where it appears. In parallel, monitor Answer Presence and Citation Share for the questions the content is meant to answer. If citation performance changes, investigate the publishing changes, content freshness, and query coverage before assigning the gain to the button.

1. Search Google's source-preferences tool for the exact domain or subdomain.
2. Choose the standard button when the CMS can safely load Google's implementation.
3. Use the supported deeplink when custom JavaScript is not possible.
4. Put the invitation near content that has earned a reader's trust.
5. Measure reader engagement separately from citation performance.

What should an AI Overview strategy look like after this update?

An AI Overview strategy should still begin with content that can answer a question accurately and show its evidence. The Preferred Sources button gives returning readers a way to identify a site they want to see more easily in Google. It does not remove the selection problem that every cited page faces: whether the page directly answers the query and gives the engine reliable material to reference.

That means editorial teams should keep building pages around the buyer's actual questions. Use clear question-shaped headings. Lead with the direct answer. Add a dated statistic, a comparison table, or original data when it genuinely supports the page. Link to primary sources rather than decorating prose with unrelated citations. These choices help a reader assess the work and give AI systems clearer material to cite.

Freshness matters especially when the question concerns a changing product feature. This release itself shows why. The May announcement established that Preferred labels were coming to AI Overviews and AI Mode. The August announcement added a new on-page activation path. A page that describes only the old settings flow is incomplete even if its basic explanation remains true.

The better editorial posture is not to chase every interface change as a ranking trick. Publish the underlying explanation, define the mechanism precisely, and update the page when the primary documentation changes. For a brand tracking AI discovery, use citation metrics to measure whether it appears in relevant answers. For a publisher with loyal readers, use Preferred Sources to make that relationship easier to express.

Google's own numbers make the audience signal worth testing. On May 27, Google said users had selected more than 345,000 unique sources. By August 20, Google said the total had passed 600,000. The before-and-after shows adoption of the preference has was growing before the embeddable button, while the release created a simpler way for publishers to request that preference on their own pages.

Sources

Source: Google Search Central

Google's publisher documentation says selected sources are more likely to appear in Top Stories and can receive a Preferred badge in AI Mode and AI Overviews for users who selected them. It also defines domain and subdomain eligibility and the button and deeplink implementations. Google Search Central, 2026-08-20

Source: Google

Google announced on May 27, 2026 that Preferred Sources were coming to AI Overviews and AI Mode, reported more than 345,000 selected unique sources, and said people were twice as likely to click through to a Preferred Source. Google, 2026-05-27

Source: Google

Google announced the embeddable Preferred Sources button on August 20, 2026 and said people had selected more than 600,000 unique sources. Google, 2026-08-20