Skip to content
GeoSalience
exp-002RunningAI-crawler hits

Does FAQPage schema raise AI crawl frequency or citation?

Half the GEO industry says FAQPage JSON-LD wins AI citations; the other half says LLMs ignore hidden markup. We add a real FAQ to one foundations explainer, hold a comparable one unchanged, and watch AI-crawler re-fetches. Running 2026-06-08; review 2026-07-08.

Hypothesis
Adding a visible FAQ section plus FAQPage JSON-LD to an article raises its AI-crawler hit frequency and (site-wide) citation rate, relative to a comparable article left unchanged.
What changed
Treatment article gains a real, visible FAQ section (drawn from its own existing content) — the FAQPage JSON-LD is then emitted automatically by the lib/seo builder. Control article is left exactly as is. No other change to either page in the window.
Treatment
Control
Metric
AI-crawler hits
Window
8 June 2026 → 8 July 2026

Status: running since 2026-06-08; review 2026-07-08. The editor selected and paired this experiment. The treatment page now carries a real, visible FAQ; the control page is held byte-for-byte unchanged for the window.

The question

Does shipping a visible FAQ with FAQPage JSON-LD actually earn more AI attention, or do the crawlers just read the page and ignore the markup? This is the smaller, on-site version of the MVP-4 Schema.org A/B test design — one matched pair on our own domain, as a first signal before any multi-domain study.

The design

  • Treatment: a real, visible FAQ section was added to GEO vs AEO vs LLMO vs SGE, built from the article's own term-comparison content. The FAQPage JSON-LD is emitted automatically by the lib/seo builder — the schema mirrors exactly the questions a reader sees.
  • Control: Knowledge cutoff and web access is left unchanged.
  • Primary metric: AI-crawler hits per page, before vs after, treatment minus control (difference-in-differences). This is the free, page-level signal from the nginx-log pipeline.
  • Window: the before window is whatever per-page crawl history exists (short — tracking began ~2026-05-31); the review is 2026-07-08.

Why these two pages

Both are foundations explainer spokes, so page type is held roughly constant — the cleaner design the original proposal said to wait for. The treatment page had FAQ-able term-comparison content but no FAQ block yet; the control is the nearest comparable foundations explainer. Residual differences in length and topic remain and are disclosed above.

The parked half: site-wide citation

The secondary signal — whether site-wide citation rate moves — needs the LLM-API citation harness, which is parked. No harness was run for this launch. If/when API spend is re-enabled, the site-wide citation trend will be attached at review as directional context only (it is never page-attributable). The primary read is the FREE crawler re-fetch frequency, already accruing.

What would count as a result

A clear, repeated rise in crawler hits on the treatment page that the control page does not share, sustained across the window. Anything smaller, on a two-page sample, we will report as inconclusive rather than dress up as a finding — and on this short before-window, an early inconclusive is the expected, honest outcome.

Analysis readout

Inconclusive

No defensible read yet for exp-002: Too few crawler hits on treatment pages (12 across both windows; need ≥ 20).

Page setBeforeAfterΔ
treatment57+40%
control56+20%
Caveats (2)
  • Quasi-experiment: a before/after observation on a small page corpus, not a randomized controlled trial.
  • AI-crawler and citation signals are noisy and lag the change by days to weeks; a short window can mislead.

Limitations

  • Single domain, single matched pair (n=2 pages). This detects only a large effect; a small one is invisible at this sample.
  • Both pages are same-type (foundations explainer spokes), which holds page type roughly constant — an improvement on the original cross-type proposal — but they are not identical in length or topic, so residual differences remain.
  • The before-window is short: per-page AI-crawler tracking only began ~2026-05-31. Per-page hits are now attributed exactly, from full per-path daily counts for the experiment's watchlisted pages (not just a crawler's daily top-10), so a low-traffic page is no longer undercounted to zero. The remaining limit is days, not attribution: a short before-window can still read inconclusive simply for lack of data.
  • Citation effect is not page-attributable: the citation harness measures the whole site, so any citation movement is context, not proof about the treatment page. Crawl frequency is the primary, page-level, FREE metric here. Site-wide citation context is DEFERRED until LLM-API spend is re-enabled — no harness was run for this launch.

Pages in this experiment

Changelog

  • Published — 31 May 2026

Raw markdown: /lab/experiments/exp-002.md