The GEO Pulse — Week 1: Perplexity Rewards FAQ Schema

Welcome to the first edition of The GEO Pulse — a weekly signal brief tracking what changed in AI search this week, who won, and what you should do about it.

This week's dominant story: Perplexity's citation engine shifted on June 9, and brands that had already implemented FAQPage structured data woke up to a citation windfall.

What Happened

On June 9, 2026, Perplexity deployed an update to its answer synthesis layer that changed how it weights source candidates when constructing responses to conversational queries. The specific change — confirmed through our longitudinal prompt tracking dataset — is an elevated preference for pages with FAQPage and Question/Answer schema markup.

The update wasn't announced. There was no changelog, no blog post. We detected it by watching citation rates across 1,840 tracked brand-query pairs shift materially within a 36-hour window.

The Two Groups

We split our tracked dataset into two cohorts: brands that had FAQPage schema on at least one high-intent page, and brands that didn't. The divergence after June 9 was immediate and consistent.

Cohort A — FAQPage schema present (n=214 brands):

• Average citation rate pre-update: 38.2%
• Average citation rate post-update: 50.0%
• Delta: +11.8 percentage points

Cohort B — No FAQ schema (n=489 brands):

• Average citation rate pre-update: 31.4%
• Average citation rate post-update: 29.1%
• Delta: −2.3 percentage points

That's not just one cohort going up. It's one going up while the other goes down. Zero-sum dynamics are in play: the citations that moved toward schema-rich pages had to come from somewhere.

Why Perplexity Is Different

ChatGPT and Claude both respond to structured data, but their citation pipelines are less sensitive to on-page schema signals — they index semantic meaning more than markup. Perplexity, by contrast, relies heavily on its live web retrieval layer, which means traditional SEO signals (including schema) carry more direct weight in its citation decisions.

This makes Perplexity a unique leverage point in GEO strategy: it's the engine where technical on-page work has the shortest feedback loop. Changes can produce measurable citation shifts within days, not weeks.

Which Query Types Were Affected Most

Not all queries moved equally. The lift was concentrated in question-format queries — "what is the best X for Y", "how does X work", "which X should I choose" — which map directly to FAQ-style content structures.

Informational branded queries (e.g., "[Brand] pricing", "[Brand] vs [Competitor]") showed a smaller but still positive lift of approximately +8% for schema-present brands.

Navigational queries ("go to [Brand] website") showed no meaningful change, as expected.

What To Do This Week

1. Audit your FAQ schema coverage. Run your top 20 pages through Google's Rich Results Test and confirm FAQPage markup is valid and crawlable.
2. Prioritize high-intent pages. Focus on product comparison pages, "how it works" pages, and any page targeting question-format queries.
3. Check your Perplexity citation baseline now. If you don't have a pre-June-9 baseline, capture your current numbers immediately — you'll want the comparison data in two weeks.
4. Don't over-engineer the schema. Four to six well-formed Q&A pairs per page outperformed pages with 20+ pairs in our dataset. Density ≠ quality.

Next Week

We're watching Claude's citation behavior around long-form content. Early signals suggest a preference shift toward posts exceeding 1,200 words with clear header hierarchies. More data needed — but it's on our radar.

The GEO Pulse is compiled from Genlytic's longitudinal prompt tracking dataset covering 2,100+ brand-query pairs across GPT-4o, Perplexity, Claude 3.7, and Gemini 1.5. Data reflects aggregate, anonymized trends.