We Audited 2,305 Business Websites.
Most Are Invisible to AI Search.
No testimonials. No cherry-picking. Just the data — every number on this page comes straight from our audit database.
A 60 means a site Google can rank but AI assistants mostly skip. When a customer asks ChatGPT or Gemini "who should I call," these sites usually aren't in the answer.
- 344 sites (15%) scored below 50 — effectively invisible to AI search
- Only 23 sites (1%) scored 80 or higher
- Biggest corpus segments: HVAC (1,055), Shopify/DTC (654), roofing (122), electricians (103)
Source: Cited Digital audit database, June 2026
- 89% — no question-and-answer content (the format AI engines quote)
- 81% — missing author names, credentials, or trust signals
- 74% — missing or broken schema markup (the labels AI reads)
- 17% — actively blocking AI crawlers in robots.txt, usually by accident
These four gaps repeat across every industry we audit. None of them require a redesign — most are fixable in days, not months.
Issue rates computed from stored audit results, June 2026
We don't publish client case studies yet — so we ran the audit on ourselves, implemented our own Fix Manifest, and re-audited. Same engine, same scoring, fully documented.
- Baseline audit of citeddigital.co scored 87 on April 30, 2026
- Implemented our own report's schema and content-structure fixes
- Re-audit on May 1, 2026 scored 91
- We keep re-auditing ourselves on a schedule: 90 as of June 1, 2026
Every number verifiable in our audit history — ask and we'll show the raw records
What Low Scorers Have in Common
Sound familiar?
All statistics on this page are computed directly from the Cited Digital audit database. Individual business sites are never named without permission. Privacy policy.
Documented by Connor Whitlock, Founder · Cited Digital · Published · Updated
Where these numbers come from
Every site in the corpus was evaluated by the same audit engine, with the same five weighted categories (Content Structure 30%, Technical 25%, E-E-A-T 23%, Schema 12%, Meta 10%). The corpus is 2,305 real business websites audited in 2026 — primarily HVAC contractors, Shopify and DTC e-commerce brands, roofing contractors, veterinarians, electricians, plumbers, and dental practices.
The aggregate statistics (average score, issue rates, score distribution) are computed directly from stored audit results. The before/after case is our own website, because it's the only engagement where we can publish every detail without anonymizing anything.
What's typically dragging scores down
Across the corpus, the same gaps account for most lost points:
1. Missing FAQ content with matching FAQPage schema
89% of audited sites have no question-and-answer content at all. Question-headed sections drive a 180% increase in AI citation rate per Previsible's 2024 prompt research. Step-by-step guide.
2. Missing author bylines and trust signals
81% of sites have weak or absent E-E-A-T signals (per Google's documented guidelines): no named author, no credentials, no verifiable profiles. "By the team" is the norm — and a meaningful weakness.
3. Missing or broken schema markup
74% of sites are missing the structured-data labels AI engines read. Every content page should have at least a content type (Article, HowTo), an organization context (Organization, LocalBusiness), and a navigation context (BreadcrumbList).
4. Blocking AI crawlers in robots.txt
17% of sites in our corpus block at least one major AI crawler — often unintentionally, as a side effect of legacy SEO plugins. It's a one-line fix that decides whether AI engines can see the site at all.
5. Walls of prose with no structure
AI engines extract from structured content — headings, lists, tables — not long paragraphs. Reading-grade level matters too: grade 6-8 gets 15% more citations than grade 11+ per SE Ranking's 2024 research.
Frequently asked questions about this data
Are these client case studies?
No, and we want to be straight about that: the statistics on this page are aggregate data from automated audits we ran across 2,305 real business websites — not paid client engagements. The one before/after we publish is our own site, because we can document every step of it. As client results accumulate, we'll publish them here with permission.
Is the before/after on your own site real?
Yes — and it's fully verifiable. Baseline audit April 30, 2026: score 87. We implemented the fixes from our own report and re-audited May 1, 2026: score 91. We re-audit ourselves on an ongoing schedule (90 as of June 1, 2026 — scores move as engines and checks evolve, which is exactly why we keep measuring). Ask us and we'll show the raw audit records.
How long does a score lift take?
Schema and crawl fixes show up in audit re-scores within 1-2 weeks. AI citation lift (the actual business outcome) takes 4-12 weeks because AI engines need time to recrawl and update their citation graphs.
Can I see what the paid report looks like?
Yes — the sample report (PDF) shows the exact format and depth you'd receive.
Will my site improve like yours did?
Depends on (a) where you start (sites under 50 have more headroom), (b) how many fixes you implement (the report ranks them by impact, so even fixing the top 3 produces measurable lift), and (c) your industry's competitive AEO baseline. Run the free audit to see exactly where you stand.
Is there a guarantee?
If your $497 paid audit returns fewer than 5 specific actionable fixes, we refund every dollar. We can't guarantee a specific score lift because implementation is on you, but we can guarantee the report contains substantively actionable work to do.
Aggregate data (industry score distributions, common-issue rates, fix-difficulty rankings) is published in our 2026 State of AEO research report for anyone who wants the bigger picture.
Want to see your own before-state?
The free audit at the top of citeddigital.co/audit gives you the same baseline every site in this corpus started from. 60 seconds, no signup, no card. Then you decide whether the $497 full report (DIY implementation) or the $1,997 Fix Pack (we implement) makes sense.