Having FAQPage schema is necessary but not sufficient. Google validates the markup, stores the answer, and then chooses whether to use it based on whether the answer stands alone when extracted.
A good answer:
"A redirect chain is when URL A redirects to URL B, which redirects to URL C. Google follows up to 5 hops before giving up. Chains longer than 2 hops leak PageRank."
A bad answer:
"It means when there are multiple redirects. This can happen for various reasons. Usually you want to avoid it."
Both answer the same question ("what is a redirect chain"). One opens with the subject named, gives concrete facts, and stands alone. The other leans on "it" and "this" without antecedent. Google lifts the first. The second stays invisible.
FAQ Answer Extractability Scorer runs 8 linguistic checks on each Q+A pair on your page.
The 8 signals
- Length 40-300 characters. Too short = not enough context. Too long = Google truncates.
- First word is a concrete subject, not a pronoun or transition. Not "it," "this," "however," "so."
- Pronoun density < 6%. Every pronoun needs an antecedent the extracted chunk doesn't have.
- Fact density. At least 2 combined numbers + proper nouns. Facts lift cite-worthiness.
- Question subject restated. If the Q contains "redirect chain," the A should too.
- Declarative opener. "A redirect chain is…" not "To understand this…"
- Direct-answer verb in first sentence. "is," "means," "describes," "contains," "provides."
- First sentence self-contained. No leading "this" / "that" without context.
Where the tool finds your FAQ content
Three source patterns:
- FAQPage JSON-LD (Question / acceptedAnswer). Gold-standard source.
<details><summary>Q</summary>A</details>blocks. Common for accessible FAQ UI.- Heading-ending-in-? followed by paragraph(s). The lowest-tech FAQ pattern.
The scorer dedupes across all three so a schema-marked FAQ that also appears as visible HTML counts once.
What 40% average looks like
In testing the scorer on 50+ FAQ pages: typical average is 35-50%. The worst answers are always pronoun-opened and fact-light. The best answers read like they were written to be quoted — which is the correct instinct.
Related reading
- FAQ Schema Parity — markup vs visible Q&A diff
- FAQ Harvester — pulls competitor FAQs
- Passage Retrievability — same signal model at paragraph level
Fact-check notes and sources
- Google FAQPage schema: developers.google.com/search/docs/appearance/structured-data/faqpage
- Google featured snippets: developers.google.com/search/docs/appearance/featured-snippets
The $20 Dollar Agency covers FAQ optimization as a repeat deliverable. The extractability scorer is the pre-ship quality gate.
