Quietly buried inside Google's May 2026 AI Mode update is a feature that's going to reshape the next 18 months of brand-led marketing — and almost nobody is talking about it. It's called Preferred Sources, and it lets a logged-in Google user "favorite" sites they want Google's AI to lean on when generating AI Overview answers.
Google's own data: users who mark a source as preferred are twice as likely to click through to that source in AI-generated answers. The implications are bigger than they look.
What Preferred Sources actually does
When a user is logged into their Google account and viewing AI Overviews or AI Mode results, they can now mark specific sites as "preferred" — either by clicking a small icon next to a citation, or by managing the list in their Google account settings. Once marked, that site gets weighted more heavily in future AI Overview generations for that user's queries.
Google also launched a "Highly Cited" label that appears next to sources that are frequently cited across many users' searches in a topic. The two systems work together:
- Preferred Sources = personalized weighting (per-user)
- Highly Cited = aggregate authority signal (across users)
If you're cited frequently enough, you get the "Highly Cited" badge, which both increases your visibility and makes you more likely to be marked as a Preferred Source by users browsing AI answers. It's a positive feedback loop — early citations beget more citations.
Why this matters more than it sounds
For 15 years, search rankings have been the scoreboard. You ranked #1 or you didn't. Now there's a new scoreboard layered on top of the old one — and it measures something different.
The old SEO question: "Am I on page one for this keyword?"
The new question: "Am I a source the AI engine trusts enough to cite — and am I cited often enough that users start preferring me?"
The first question has a deterministic answer (you check rank tracker, see your position). The second has a behavioral one — your citation rate compounds based on how often you're surfaced, which compounds based on how often you're preferred, which compounds based on whether the AI's first-time citation of you was a positive experience for the user.
This is the first time in modern search where brand affinity directly drives algorithmic visibility. Not via backlinks. Not via brand search volume. Directly. By a user clicking "I prefer this source."
The three implications for your marketing
1. The brand layer of marketing just got more accountable
"Brand awareness" has always been hard to attribute to revenue. Now it isn't. If your brand investment makes 100 customers actively mark you as a Preferred Source in their Google account, those 100 customers are twice as likely to click through to you next time they search anything tangentially related. That's a measurable, compounding effect — the kind of thing CFOs have always wanted brand spend to be and rarely is.
Practical implication: brand-side investments (podcast sponsorships, thought leadership content, community building, conference presence) suddenly have a downstream click-through metric you can correlate against. If your brand campaigns coincide with a measurable lift in Preferred Source marks, the brand spend is paying off in a way you can defend in a budget meeting.
2. The first AI citation matters more than the tenth
The user's first encounter with your brand inside an AI answer is the one that decides whether they mark you Preferred or scroll past. If your AI-cited content is generic, surface-level, or worse — wrong — you don't just lose that one citation. You lose every future opportunity for that user to weight you heavily.
Practical implication: the editorial standard for content that's eligible to be AI-cited just went up. The "ship volume to capture more queries" playbook is dead. The "ship few pieces of exceptional content per topic" playbook is alive. Quality over throughput.
3. "Highly Cited" becomes a real scoreboard
For categories where the same 5-10 sources are consistently cited, the "Highly Cited" label is going to become a visible mark of category authority. Customers will see it. Buyers will see it. Investors will see it. It's the new "Inc. 500" or "Forbes 30 Under 30" — except it's algorithmically earned, recurringly applied, and tied to actual category usage.
Practical implication: if you can earn the "Highly Cited" badge in your category, you've earned a piece of permanent real estate in your prospects' decision-making. Worth allocating multi-quarter content investment toward.
What to do about it this quarter
Three concrete moves:
One: audit your AI citation rate today. Open ChatGPT, Perplexity, Gemini, and Google AI Mode. Run the 10 queries your customers most often ask. Count how many cite you. Count how many cite your top three competitors. The gap is your starting point.
Two: identify the two or three queries where being cited would matter most. Probably high-intent, high-commercial-value, low-competition. Map the content gap — what would Google or ChatGPT need to find on your site to consider citing you for those queries? Write those pieces. Make them undeniably the best resource for that question.
Three: structure your content so the AI citation lands well. Open with the answer. Define terms clearly. Use semantic HTML. Include data points and named examples that the AI can quote without ambiguity. The goal is for the first AI citation of your content to read like a credible authority, not a marketing pitch.
The honest part
"Highly Cited" is going to be a vanity metric for some businesses and a meaningful lever for others. If your customers don't actually use Google AI Mode, none of this matters yet. If your customers do — and the trajectory is clear that they will — the next 18 months are when the new category-authority game gets played. The brands that show up early will compound; the brands that wait will be playing catch-up against established Preferred Sources.
The reason I'm watching this closely: Google AI Mode just passed one billion monthly users. Whatever you think about AI search adoption, the numbers say it's already mainstream. The behaviors users build inside AI Mode are going to set the search defaults for the next decade.
If you want a candid read on your current AI citation rate — yours and your top 3 competitors' — across ChatGPT, Perplexity, Gemini, and Google AI Mode, that's what our free Marketing Score measures. A senior strategist runs the queries personally, scores you against your real category competitors, and emails you the result within one business day. No automation. No template. Just the actual numbers.