Reading time 13 minutes
Organic visibility no longer depends solely on ranking highly for a keyword. With AI Overviews, AI Mode and increasingly personalised results, Google is also reshaping the relationship between users, brands and the sources they consider trustworthy.
In this context, Google Preferred Sources is a feature that allows individuals to indicate which websites they want to see more frequently in certain search experiences.
For media outlets, specialised blogs and companies that regularly publish informative content, this development represents a significant opportunity. An audience that already trusts a brand can express that preference within Google, making it easier for that brand's content to stand out to those same users.
However, it is important to understand the scope of the feature: it is not a certification, it does not guarantee rankings and it does not make a domain a preferred source for the entire market.
The real opportunity is more strategic. Google Preferred Sources brings together customer loyalty, editorial authority and visibility in artificial intelligence environments. This means that SEO is no longer just about demand generation; it now has a second role: turning occasional readers into an audience that recognises, remembers and actively chooses the source.
What Google Preferred Sources is and what it changes
Google Preferred Sources is a search personalisation option. Users select domains they want to prioritise and, when those sites publish content relevant to a query, Google may show them more frequently and label them as a preferred source.
Google Search Central's official documentation on Preferred Sources states that the feature is available globally in Top Stories, in languages supported by Google Search. The selected sources may also be highlighted in AI Overviews and AI Mode wherever these experiences are available.
Control belongs to the reader. A brand cannot declare itself a preferred source or request privileged treatment from Google on a general basis. It can check whether its domain appears in the preferences tool and encourage its community to select it, but the final decision is made by each individual. Therefore, the effect is felt in the experience of those who have selected the source, rather than across all Google users.
This distinction avoids a common misconception. Preferred Sources does not replace ranking systems, topical relevance or indexing requirements. The content must still be appropriate for the query and eligible to appear. The preference reinforces an existing relationship, but it cannot compensate for a weak, outdated or technically inaccessible page.
Google has also stated that readers are twice as likely to click on a site after marking it as a preferred source. This figure was shared during the international expansion of Preferred Sources.
The figure does not mean that traffic to any given domain will double. It describes an aggregate difference in behaviour and confirms something more useful: when user trust becomes visible within the interface, the likelihood of a visit may increase.
How Google Preferred Sources drives visibility in AI
The inclusion of preferred sources in AI Overviews and AI Mode changes the strategic value of the feature. In these experiences, Google synthesises information and provides supporting links alongside the answer. When a source selected by the user is relevant, it may be highlighted with the corresponding label. The brand gains recognition in an environment where clicks no longer depend on a traditional list of blue links.
However, this advantage remains personalised and limited to the Google ecosystem. Selecting a domain as a Preferred Source does not affect its visibility in ChatGPT, Claude, Perplexity or other assistants. Nor is there any official evidence that accumulating selections in itself provides a ranking boost for users who never selected the source.
The relationship with AI is better understood by looking at how Google retrieves information. Its systems may use a query expansion technique known as query fan-out to perform related searches across subtopics and identify supporting pages.
According to Google's official guidance on AI features and websites, no additional technical requirements or special markup are needed to appear in AI Overviews or AI Mode. Pages must be indexed, be eligible to appear with snippets and meet the usual Search fundamentals.
The key is to combine two aspects. The first is eligibility: ensuring that Google can crawl, understand and retrieve the content. The second is preference: getting people to value the source enough to choose to prioritise it. A strong strategy addresses both. Publishing more content without building trust limits retention, while promoting the button without having relevant content to offer reduces the incentive to choose the brand.
From acquisition to brand preference
Traditional SEO tends to focus on reaching people who do not yet know a company. Google Preferred Sources adds a retention layer. A visit from Search, a newsletter or a social network can become a lasting signal of affinity if the user adds the domain to their preferences.
This is particularly relevant for organisations with a recognisable editorial voice: consultancies that explain regulatory changes, technology companies publishing their own research, professional firms producing practical guides, or specialist media outlets covering a particular sector. In all these cases, the value lies not only in answering an individual query, but in becoming a trusted reference that readers want to return to.
In practice, this means designing content as a relationship-building asset. The information should meet the immediate need, but also give readers a reason to remember the source: an original framework, proprietary data, a clear methodology, practical examples or an interpretation that connects the change to business decisions. The preference should be requested after demonstrating value, not before.
Technical requirements and configuration
Google establishes an important structural condition. The tool supports sites at the domain or subdomain level, such as example.com or insights.example.com, but not a subfolder like example.com/blog.
This doesn't mean that posts hosted in a folder will stop being indexed or appearing in Search. It means that the unit that can be selected as a source is the domain or subdomain recognised by the tool.
Before launching a campaign, it is advisable to search for the site on the official preferences page. If it appears, Google offers three ways to make it easier for users to select it:
- Insert the standard button using JavaScript. This is the option recommended by Google because of its integrated and localised experience.
- Use the advanced implementation. This lets you customise the design and when the button is activated.
- Create a direct link. This is a useful alternative when the CMS or technical environment does not allow the button to be loaded.
The standard implementation requires loading the official library and adding the button container to the page. The direct link can also be used in newsletters, social media or campaigns targeting an existing community. In all cases, the CTA should explain the benefit to the reader, for example, receiving more content from the source in relevant Google search results.
The technical fundamentals must also be maintained. robots.txt, the CDN and server must allow Googlebot to crawl the site; the relevant pages need internal links; the main content must be available as text; structured data must match what is visible; and the page experience must be consistent.
Google clarifies that no specific AI file or special schema is required for these experiences. The technical and editorial fundamentals of SEO remain the foundation of visibility.
Editorial strategy for becoming a preferred source
The button is the final step in a strategy, not its starting point. For someone to select a source, they must first recognise a consistent and differentiated value proposition. An effective editorial plan can be built around five decisions.
- Define a recognisable topical territory. The brand should know which subjects it aims to be consulted on and what perspective it can sustain over time. Covering too many topics without sufficient depth dilutes the association between the source and its area of expertise.
- Create non-interchangeable information. Original studies, data analysis, documented case studies, interviews, transparent comparisons and decision-making frameworks provide more reasons to be cited and remembered than a generic summary of what has already been published.
- Build recurring formats. Regular reports, observatories, updated guides and analyses of the latest developments help create expectations. Readers know what they will receive and why it is worth keeping the source close at hand.
- Place the CTA at moments of trust. The end of a useful article, subscription confirmation or a newsletter with a high level of engagement offer more natural opportunities than an intrusive pop-up when entering the site. The message should be specific, understated and consistent with the user experience.
- Distribute the invitation among your existing audience. Email, communities, social media and customer areas can engage users who already know the brand. Asking cold traffic to select the source generally makes less sense than doing so after a sign of interest.
This approach also improves the likelihood of appearing in generative experiences through channels that do not depend on the preference. In its guidance on optimising content for generative AI features, Google recommends creating valuable, original information designed for people. It also cautions against mass-producing pages designed to capture every possible variation of a query.
The upshot is clear: Sustainable authority is built on information that deserves to be retrieved, linked to and chosen.
How to measure the impact without misleading attribution
Measurement requires caution because clicking a button on your own site does not necessarily demonstrate that the user has completed the selection. If the implementation allows the interaction to be recorded, that event should be interpreted as an indication of intent or the start of the process, not as a confirmed preference, unless Google provides a verifiable signal.
The analysis should combine four layers: organic impressions and clicks, visibility in generative features, recurring audience behaviour, and business outcomes.
Search Console is gradually rolling out specific reports on performance in generative AI features. These reports allow users to view impressions, URLs, countries, devices and changes over time in AI Overviews, AI Mode and certain Discover experiences.
It is advisable to establish a baseline before activating the CTA and compare equivalent periods. You can monitor recurring-user traffic, conversions assisted by content, growth in branded searches, the performance of pages where the button is displayed, and changes in the URLs visible in AI.
Where possible, a controlled test by section or audience will help distinguish the effect of the CTA from editorial, seasonal or algorithmic changes.
No single metric proves causality. An increase in impressions may result from broader topical coverage; an increase in clicks may result from better titles; and an improvement in conversions may result from a clearer offering. Useful evaluation connects exposure, preference, visits and commercial outcomes, while remaining cautious when attributing each change.
Errors that can reduce the value of the initiative
The first error is presenting Google Preferred Sources as a button to “appear first in AI”. This promise confuses personalisation with ranking and can undermine credibility.
The second is deploying the CTA aggressively before the user has received any value. The interruption contradicts precisely the relationship of trust that the initiative aims to strengthen.
It is also risky to create large volumes of superficial content to cover every possible AI question. Google states that its generative features rely on the core Search systems and discourages inauthentic mentions, unnecessary markup and tactics designed to manipulate responses.
Automation can help with research, structuring or updating, but publication requires expert judgement, verification and a recognisable contribution.
Finally, Preferred Sources should not be confused with Google Cloud partner programmes, Marketplace presence, or inclusion in models and agents. These are separate mechanisms, with independent objectives and processes. An AI company can develop a preferred editorial voice while simultaneously building a technological relationship with Google, but one route does not automatically grant access to the other.
Google Preferred Sources as a trusted asset
Google Preferred Sources does not eliminate the uncertainty surrounding rankings or provide an immediate formula for recovering traffic absorbed by generative answers. Its value lies elsewhere: it turns editorial trust into an explicit preference within Google and allows brands to strengthen their relationship with people who already recognise their value.
The evolution of the feature confirms that Google wants to strengthen this relationship between audiences and sources. In August 2026, the company introduced a new interactive Preferred Sources button for publishers, designed to allow readers to add a publication and return to the content they were viewing later.
For ATLS Global, the strategic takeaway is clear. First, it must ensure indexability, relevance and differentiated quality. The organisation can then integrate the button or direct link into a well-designed audience experience, engage its own communities and measure the entire journey using realistic criteria.
Visibility in AI is not achieved through an isolated label, but by building a source that Google can retrieve and that users want to choose.
FAQs about Google Preferred Sources
What is Google Preferred Sources?
Google Preferred Sources is a personalisation feature that allows users to choose sites they want to see more frequently in Top Stories and, where available, in AI Overviews and AI Mode.
Does Google Preferred Sources improve SEO rankings?
Google Preferred Sources can increase a site's visibility for users who have selected it, but it does not guarantee better overall rankings and does not replace relevance, indexing or content quality.
How can a company appear in Google Preferred Sources?
It should check whether its domain or subdomain appears in the official tool. It can then facilitate selection using Google's JavaScript button or a direct link aimed at its audience.
Does Google Preferred Sources help a site appear in AI responses?
Yes. A selected source may be highlighted with the Preferred label in AI Overviews and AI Mode for that user, provided that the content is relevant and eligible to be shown.
How is Google Preferred Sources measured?
It is recommended to combine CTA interactions, recurring traffic, conversions, branded searches and AI impressions in Search Console, while avoiding treating every click as a confirmed selection.
