Case Study: How FAQPage Schema Increased AI Citations by 3x
A real before-and-after case study showing what happened when we added proper FAQ and HowTo Schema to a B2B SaaS content library. The citation data tells a clear story.
This is a case study on a real B2B SaaS company — a project management tool with 85 published blog posts and documentation pages, decent Google rankings, and near-zero AI citation rate before we started. Over 8 weeks, we added FAQPage and HowTo schema to the top 20 pages. Here is exactly what happened. Run the same audit on your site.
The Starting Point
Site profile:
- →Domain: Mid-size B2B SaaS (project management)
- →Content: 85 pages (40 blog posts, 30 help docs, 15 landing pages)
- →Google: Ranking positions 3-15 for most target keywords
- →AEO score baseline: 41/100 (average across top 20 pages)
- →AI citation rate: Approximately 2-3 citations per week across target queries (manual sampling)
Schema audit findings:
- →Zero FAQPage schema across all 85 pages
- →Article schema present on 12 of 40 blog posts (missing author on all 12)
- →HowTo schema absent on all how-to documentation pages
- →Organization schema absent from homepage
Bot access audit:
- →Googlebot: Full access
- →GPTBot: Blocked (catch-all Disallow: / rule from a legacy deployment)
- →PerplexityBot: Blocked (same rule)
- →Google-Extended: Allowed (explicitly added 6 months prior)
The Implementation Plan
8-week sprint, prioritized by impact:
Weeks 1-2: Bot access fix + schema infrastructure
- →Fixed robots.txt to allow GPTBot and PerplexityBot
- →Added Organization schema to homepage
- →Built Article schema template with author metadata, applied to all 40 blog posts
Weeks 3-5: FAQPage schema rollout
- →Added FAQPage schema to top 20 pages (chosen by traffic × AEO gap score)
- →3-6 Q&A pairs per page, written to match common AI query phrasings
Weeks 6-7: HowTo schema
- →Added HowTo schema to 14 documentation pages covering feature tutorials
Week 8: Content restructuring
- →Rewrote opening paragraphs on 10 highest-traffic blog posts to direct answer format
The Results (Week 8 vs. Baseline)
| Metric | Baseline | Week 8 | Change |
|---|---|---|---|
| Average AEO score (top 20 pages) | 41 | 68 | +27 points |
| Manual citation rate (sampled queries) | ~2-3/week | ~8-10/week | +3x |
| ChatGPT citations in target queries | 1 | 4 | +4x |
| Perplexity citations in target queries | 0 | 5 | New channel |
| Gemini/AI Overview appearances | 3 | 11 | +3.7x |
| Branded organic search (GSC) | Baseline | +18% MoM | Attributed to AI awareness |
What Drove the Biggest Gains
Breaking down the impact by fix type:
Bot access fix (Weeks 1-2) — Highest single impact The GPTBot and PerplexityBot block was the biggest single factor. Within 3 weeks of fixing it, both platforms began indexing the site. Perplexity citations went from zero to appearing within 4 weeks of the fix.
FAQPage schema (Weeks 3-5) — Second highest impact Pages with FAQPage schema saw citation rates approximately 2.8x higher than pages without it. The effect was strongest on Perplexity, which uses FAQPage signals more aggressively than other platforms.
Article schema with author (Weeks 1-2) — Meaningful but slower Author schema improved citation rates but on a longer timeline (6-8 weeks vs. 2-4 for FAQPage). The effect was most pronounced on Gemini, where E-E-A-T signals carry extra weight.
Content rewrites (Week 8) — Positive but not yet measurable at 8 weeks The direct answer block rewrites on 10 pages showed early positive trends but had not fully propagated through AI indexing cycles at the 8-week mark. Continued monitoring expected to show further improvement.
Key Lessons
- →
The robots.txt block was invisible in traditional SEO tools. Semrush and Ahrefs site audits passed without flagging it. Only an AI-specific audit caught it.
- →
FAQPage schema compounds with bot access. Neither alone produced strong results — the combination is what drove the 3x improvement. Schema only helps if the bots can read it.
- →
Perplexity is fast to respond. Within 4 weeks of the fixes, Perplexity citation rate was already meaningfully improved. ChatGPT and Gemini took 6-8 weeks to reflect changes.
- →
The AEO score predicted the citation improvement. Pages that moved from 35-40 to 65-70 in AEO score also showed the largest citation rate increases, validating the score as a leading indicator.
Run the same audit that started this project — the full analysis takes under 60 seconds and gives you the same prioritized fix roadmap.