Google Preferred Sources: why you cannot optimise your way past a brand the reader has already chosen

You can optimise title tags, fix Core Web Vitals and build links for months, and still lose to a company the reader already knows. Google Preferred Sources makes an uncomfortable truth plain to see: it is no longer just technical optimisation that decides what gets seen – the brand does too.

What is Google Preferred Sources?

Preferred Sources lets users choose which publishers they want to see more often in Google's Top Stories. On 30 April 2026 the feature was extended to all languages, and on 27 May 2026 to Google's AI Overviews and AI Mode. So a source you have chosen also gets a visible place inside AI-generated answers, not just in the traditional list of search results.

According to Google, users have so far chosen more than 345,000 unique sources, and readers are roughly twice as likely to click on a link from a source they have selected. Search Engine Journal describes the feature as a user-declared signal that works alongside Google's ranking systems. Google itself is more cautious, saying only that the content becomes "more likely to appear in Top Stories".

Why is this bigger than a new feature?

For ten years, Google's line was that brand is not a direct ranking factor. John Mueller confirmed in 2016 that Google has no "website authority score". The logic behind Preferred Sources is not new; what is new is that Google now shows the mechanism openly to users.

This is not an isolated piece of news but part of a pattern. The user's preference becomes an explicit signal in traditional search, and the same preference shapes which sources are highlighted in AI answers. Both layers of search are moving in the same direction: towards publishers that people recognise and actively choose.

Does brand really matter more than links?

The data suggests so. In an analysis of 75,000 brands by Ahrefs, brand mentions correlated at 0.664 with visibility in AI Overviews. Backlinks correlated at 0.218 and domain authority at 0.326. In other words, how often a brand is mentioned around the web was more strongly linked to AI visibility than either links or domain strength. The table below shows the full ranking of the factors Ahrefs measured.

An important caveat: this is correlation, not proven causation. Ahrefs used Spearman correlation and only included domains with a domain authority above 40. But the direction is clear and shows up in several measurements. We cover how this plays out in AI answers specifically in AI Overviews and GEO.

Why rank tracking reports on the wrong layer

If brand recognition drives both traditional visibility and AI answers, average search position becomes a blunt metric. Rank tracking tells you what happened, but less and less about what drives results. You can rank first for a niche term and still be invisible in the AI answer where the deal actually begins.

Rank tracking is not worthless. It is still good for finding technical problems and comparing individual pages. But as the primary KPI reported to management or a client, it measures the wrong thing. Better proxy metrics are brand mention volume, share of voice in AI answers and how often the brand is mentioned in relevant contexts.

What do you do if you do not have 345,000 fans?

Large brands have a head start here, and no amount of technical optimisation will close the gap. But brand signals are built, not inherited, and for a company with 5 to 50 employees there are concrete levers:

• Founder content: a recognisable sender with a face and an opinion is mentioned and shared more often than an anonymous company account.
• Earned AEO and PR: mentions in trade media, podcasts and newsletters are now an SEO investment, not just PR.
• A consistent entity profile: the same name, description and facts everywhere (your own website, LinkedIn, Wikidata, industry directories) make it easy for both Google and AI to connect the mentions to one entity.
• Content worth citing: data, clear answers and your own positions get mentioned; generic content goes unnoticed.

You cannot buy a place as someone's preferred source. But you can make the brand worth choosing – a slower but more durable investment than chasing rankings.

What does it mean for your SEO strategy in 2026?

Technical SEO and on-page optimisation are not going away. But they are the floor, not the ceiling: they build on brand recognition rather than replacing it. Anyone who sees SEO as purely technical will lose ground to those who also build a brand worth remembering.

Strong SEO and AEO come from the combination: strategy, copy that both people and AI want to cite, business understanding and a technical foundation. Technology alone produces no citable answers; copy alone does not give machines the structure they need. Would you like to know where your brand stands in that equation? Get in touch.