Before & After: How Adding FAQ Schema Increased AI Citations by 3x
A real case study showing what happened when we added proper FAQ and HowTo Schema to a client's website. The citation data before and after tells the full story.
Background
In late 2024, a B2B SaaS company in the project management space came to us with a familiar problem: solid Google rankings, decent traffic, but essentially zero visibility in AI-generated answers. Their product was being discussed in ChatGPT and Perplexity responses regularly — but usually as a passing mention rather than a cited source.
Their site had 47 pages of editorial content — blog posts, guides, and comparison articles. All were well-written and comprehensive. None had any Schema markup beyond a basic Organization Schema on the homepage.
Client profile (anonymized)
DA 41
Domain Authority
47
Content pages
~8,200/mo
Organic visits
The baseline audit
We ran a full RankAsAnswer audit across all 47 pages. The results confirmed the pattern we've seen across hundreds of similar sites:
Before: AEO Audit
What we found
- !31 pages had FAQ-style content with no FAQPage Schema
- !12 pages had step-by-step guides with no HowTo Schema
- !All blog posts had bylines but no Article Schema
- !Most H2s were declarative, not question-phrased
What we changed
We focused on the 15 highest-traffic pages first. For each page, we made three changes:
Added FAQPage Schema to all pages with Q&A content
RankAsAnswer generated the JSON-LD for each page automatically. We reviewed and pasted the code into each page's head section. This took approximately 90 minutes for 15 pages.
Added Article Schema with author attribution
For all 15 blog posts, we added Article Schema including the author's name, sameAs links to their LinkedIn and Twitter, and both datePublished and dateModified fields.
Rewrote the most important H2s as questions
On each page, we identified 2-3 key H2 sections that answered specific user questions and rephrased them as questions. This was a content edit, not a rewrite.
Here's what the before and after looked like for a typical page:
<!-- No structured data found on this page -->
<!-- Only visible content: Q&A section in plain HTML -->
<div class="faq-section">
<h3>What is Schema markup?</h3>
<p>Schema markup is a vocabulary of structured data...</p>
<h3>How do I add Schema to my site?</h3>
<p>You add Schema using JSON-LD in a script tag...</p>
</div><script type="application/ld+json">
{
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "What is Schema markup?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Schema markup is a vocabulary of structured data added to HTML that helps machines understand the meaning and type of your content."
}
},
{
"@type": "Question",
"name": "How do I add Schema to my site?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Add a script tag with type 'application/ld+json' containing your JSON-LD Schema in your page's head or body."
}
}
]
}
</script>Results after 6 weeks
After: 6 Weeks Post-Implementation
Unchanged metrics
The traditional SEO metrics didn't change — only the AEO signals did, confirming that the citation improvement was caused by Schema, not other factors.
Why it worked
The improvement was almost entirely driven by FAQPage Schema. The content already existed — the Q&A pairs were in the HTML. Adding Schema didn't change the content; it made the content machine-readable in a way that AI retrieval systems could extract and cite directly.
Perplexity's Sonar model in particular showed the fastest response — within 3 weeks of the Schema additions, the pages began appearing in Perplexity's Sources panel for relevant queries.
Total implementation time