The Situation
We recently completed an AI Visibility Intelligence Report for a five-year-old marketplace platform with genuine product-market fit and billion-dollar ambitions. Not a startup finding its feet — a real business, with a real user base, actively targeting the kind of growth trajectory that ends in a valuation conversation. The kind of company where the product isn't the problem.
The visibility infrastructure is.
Being readable by AI is not the same as being recommended by AI. This platform scored a perfect 3/3 on crawlability — every page fully accessible to every AI engine. It scored 0.5/4 on citation readiness. Not one AI engine cited it, recommended it, or classified it consistently. Crawlability got them in the room. Schema and positioning decide whether they get cited.
What the Scan Found
The AI Visibility Intelligence Report covers 17 signal checks across six AI engines: ChatGPT, Gemini, Grok, Perplexity, Brave, and Claude. Here is what it found for this client.
| Signal Area | Score | Finding |
|---|---|---|
| Crawlability | 3 / 3 | Perfect. Every page fully readable by all AI engines. No technical barriers. |
| Structured Data | 1 / 2 | No JSON-LD Organization schema. AI has no machine-readable answer to "what is this company?" |
| AI Hygiene | 0.5 / 1.5 | Weak signals. Inconsistent brand descriptions across pages and platforms. |
| Citation Readiness | 0.5 / 4 | Not cited by ChatGPT, Gemini, or Brave. Zero brand mentions on X in 7 days. |
| Cross-Engine Classification | 0 / 6 | All 6 engines classified the company differently. None placed it in a stable category. |
| Overall Score | 4.5 / 10 | "Partially Visible" — readable but not recommendable. |
The Root Cause
The client built a real product. The problem wasn't the product — it was that the company had never decided, in terms AI can process, what it is.
No Organisation schema. Inconsistent positioning signals across pages. No authoritative third-party mentions to anchor the brand in AI training and retrieval systems. The result: six AI engines, each doing their best to infer a classification from conflicting signals, each arriving at a different answer.
AI engines aren't being difficult. They're doing exactly what they're built to do: synthesise available signals into a confident recommendation. When those signals conflict or are absent, the engine defaults to silence — or worse, to a competitor whose signals are clear.
What We'd Fix
Deploy JSON-LD Organization Schema
Give every AI engine a single, authoritative, machine-readable answer to "what is this company and what does it do." This is the fastest single fix with the highest downstream impact on classification consistency.
Unify Brand Positioning Across Every Surface
Audit every page — homepage, about, product, meta descriptions — for consistent category language. If different pages say different things, AI engines average them into confusion. One clear, repeatable description of what the company is and does, used consistently everywhere.
Build Citation Infrastructure
Structured outreach to relevant publications, directories, and industry communities to generate third-party brand mentions. AI systems use external validation signals to confirm classifications. Without them, even a perfectly-schemed site remains unanchored.
The Lesson
Perfect crawlability is table stakes. It means AI can see you. It says nothing about whether AI will recommend you.
Recommendation depends on schema clarity, positioning consistency, and external validation. A 3/3 crawlability score combined with a 0.5/4 citation readiness score isn't a contradiction — it's the most common pattern we see. Companies invest heavily in being readable and almost nothing in being understood.
Most established companies have this problem. The brands that win in AI-generated answers are the ones that made it easy for AI to be confident about them.