Learn · Web Foundations
What AI Overviews Actually Reward
Classic SEO optimized for blue links. AI Overviews optimize for synthesis. The signals that determine whether ChatGPT, Perplexity, and Google AI cite your brand are not the signals your old SEO playbook trained you to chase.
AI Overviews reward a different signal set than classic SEO. Brand mentions across editorial properties, citation density, structured data, and semantic coherence now matter more than raw backlink counts — and the brands winning in AI answers are the ones who treated this as a strategy shift, not a tactic.
The thesis
- Why backlinks dropped from ~40% to ~10% of citation weight in AI engines
- The five core signals that determine AI Overview inclusion
- How to audit your own pages against the AEO Signal Set
- Which schemas (Article, FAQPage, HowTo, Organization) AI engines actually parse
- A 12-month editorial cadence that hits every AEO signal monthly
The framework: The AEO Signal Set
Answer Engine Optimization is not SEO with a new acronym. It is a different optimization function, scoring a different set of signals. The AEO Signal Set names the five inputs that AI engines weight most heavily when deciding which brands to cite.
Brand mention density across editorial properties
AI engines lean heavily on entity co-occurrence. When your brand is mentioned alongside category-defining terms across reputable editorial sites — not just on your own domain — language models learn the association. This is the single largest signal shift from classic SEO, where on-domain optimization dominated.
Structured data coverage
Article, FAQPage, HowTo, Organization, and Person schemas give AI engines machine-readable confirmation of what your page is, who wrote it, and when. Pages without structured data are guessable. Pages with comprehensive schema are parseable — and parseable pages get cited more often.
Entity recognition and semantic coherence
AI engines need to know that your brand is a distinct entity, linked to the right Wikipedia page, Wikidata ID, and sameAs identifiers. They also reward pages where the topic, examples, and supporting facts form a coherent semantic neighborhood. Ambiguous brands and thin pages are routinely passed over for synthesis.
Original research and proprietary data
AI Overviews preferentially cite sources that contribute novel facts, statistics, or frameworks — not commentary on someone else's data. If your page contains a number nobody else has, that number becomes a citation magnet. This is why named frameworks and proprietary datasets outperform generic thought leadership in AI citation.
Author authority and credentialing
Author schema with credentials, sameAs links to author profiles, and consistent authorship across an editorial network all contribute to author-level trust scores. AI engines are increasingly modeling who wrote a page, not just what the page says — and named, credentialed authors carry more citation weight than anonymous staff posts.
The data.
Why the signal set changed
Classic search engines were ranking systems. Given a query, they returned a list of ten candidate URLs ordered by likelihood of relevance. Backlinks worked well in that world because they were a strong proxy for authority — pages other sites pointed to were demonstrably useful. The user did the synthesis after the click.
AI engines are synthesis systems. Given a query, they generate an answer and then decide which sources to cite as evidence for that answer. The optimization function changed from 'which page is most useful' to 'which source best supports this specific claim.' That shift quietly demoted backlinks (a popularity signal) and promoted entity clarity, schema, and semantic coherence (synthesis-friendly signals).
The brands winning in AI Overviews are not necessarily the ones with the largest backlink profiles. They are the ones whose pages are easiest for a language model to read, verify, and quote. That is a different optimization problem, and it requires a different content strategy.
How AI engines actually pick citations
When an AI engine generates an answer, it runs a retrieval step (finding candidate documents), a synthesis step (generating the response), and a citation step (attaching sources to claims). Most SEO professionals optimize for retrieval — the same intuition as ranking. The brands winning AEO optimize for the citation step.
Citation selection rewards specificity. If your page contains the exact statistic, framework name, or quoted definition the AI engine needs to support its claim, you get cited. If your page restates common knowledge in generic language, the engine has dozens of equivalent sources to choose from — and you become interchangeable. Original research, named frameworks, and proprietary data are not branding flourishes. They are citation hooks.
This is why brands publishing through Pillar Studio consistently outperform brands publishing only on their own domain. Editorial network distribution multiplies the surface area where your brand can be cited, while named frameworks and proprietary data give engines a reason to cite you specifically rather than a competitor.
The structured data layer
Schema.org markup is not new, but its role has shifted. In classic SEO, schema was a 'nice to have' that occasionally produced rich snippets. In AEO, schema is the machine-readable confirmation that lets engines parse your page with confidence. Pages without schema are guessable; pages with schema are parseable; parseable pages get cited.
The four schemas that matter most for AEO are Article (with author, datePublished, publisher, and image), FAQPage (which AI engines parse aggressively for Q&A-shaped queries), HowTo (for procedural content), and Organization (with sameAs links to Wikidata, Wikipedia, LinkedIn, and Crunchbase). Author schema, linked to a Person entity with credentials, adds an additional trust layer that AI engines are increasingly weighting.
If you do nothing else after reading this piece, audit your top ten commercial pages for these four schemas. The gap between the average brand's schema coverage and the brands routinely cited in AI Overviews is usually larger than expected — and easier to close.
Why brand mentions now matter more than backlinks
Language models learn associations from co-occurrence. When your brand name appears alongside category terms in editorial content across reputable sites — even without a hyperlink — the model learns that you belong in that category. This is the technical reason 'unlinked brand mentions' have become a measurable AEO signal.
Backlinks still matter, but the marginal return has compressed. A backlink from a generic blog adds less than a brand mention in a category-defining editorial piece on a high-authority publication. This is why Pillar Authority structures placements around editorial integration rather than link-building — the citation value of being named in the right context exceeds the link equity of being linked from a generic one.
The implication for budget allocation is significant. Brands still spending the majority of their off-domain budget on link-building are optimizing for an algorithm that has moved on. Brands shifting toward editorial network mentions, authored thought leadership, and schema-rich on-domain content are building the citation surface AI engines actually reward.
Apply this to your work
Run this audit against your top commercial pages and your competitors' citation patterns. Most of these are work you can do this week.
- Search your top 3 commercial queries in ChatGPT, Perplexity, and Google AI Overviews. Record which sources are cited and what schema those sources use.
- Audit your top 10 pages for Article schema with author, datePublished, publisher, and image fields populated.
- Add FAQPage schema to any page with visible Q&A content — mirror the visible questions exactly.
- Verify your Organization schema includes sameAs links to Wikipedia, Wikidata, LinkedIn, Crunchbase, and your verified social profiles.
- Identify at least one proprietary data point, statistic, or named framework on each cornerstone page. If you do not have one, commission original research.
- Map your top 20 brand mentions across editorial properties in the past 12 months. Identify three high-authority publications where you are absent but competitors are present.
- Build a 12-month editorial cadence that hits each AEO signal monthly: original data, named framework, editorial mention, schema audit, author authority piece.
Where this connects to Pillar
The AEO Signal Set is the operating model behind every Pillar offering. Pillar Studio builds the on-domain editorial assets and schema architecture that make your pages parseable. Pillar Authority places your brand and named frameworks into the editorial network where AI engines learn category associations. Pillar Institute trains your team to maintain the signal set in-house.
Frequently asked questions.
Do backlinks still matter for AI Overviews?
Yes, but less than they did. Our estimate is that backlinks have moved from roughly 40% of citation weight to roughly 10% over the past three years. They remain a useful signal — particularly from high-authority editorial properties — but the marginal return on a generic link has compressed sharply. Brands still spending the majority of off-domain budget on link-building are optimizing for the wrong era of the algorithm.
Which schema should I prioritize if I can only implement one?
Article schema with author, datePublished, publisher, and image fields populated. It is the single most universally parsed schema across AI engines, and it unlocks author authority signals as a downstream benefit. After Article, prioritize Organization schema with sameAs links, then FAQPage for any page with visible Q&A content.
How do I get unlinked brand mentions on editorial properties?
Through earned editorial placement — contributed thought leadership, expert sourcing for journalists, named framework distribution, and original research. This is the core work of Pillar Authority. The brands cited frequently in AI Overviews tend to have a steady cadence of editorial mentions across a defined network of high-authority publications, not a single PR splash.
How quickly do AEO signals show up in AI citations?
Faster than classic SEO, but not instantly. Structured data changes are typically reflected within days to weeks as engines recrawl. Brand mention density and entity recognition shifts take longer — typically three to six months of consistent editorial cadence before AI engines re-weight your brand within a category. Original research and named frameworks can produce citation pickup within weeks if distributed correctly.
Is AEO replacing SEO, or supplementing it?
Supplementing it, for now. Classic SEO traffic still exists and still matters, but the share of queries answered directly by AI Overviews is growing — already past 40% of US queries — and the click-through rate on classic results below an AI Overview is meaningfully lower. The brands that will compound advantage over the next 24 months are the ones treating AEO as a parallel discipline with its own signal set, not a footnote to their SEO strategy.