50
| Retrievability | 100 |
| Claim density | 64 |
| Structure | 41 |
| Freshness | 0 |
| Schema validity | 0 |
| Entity consistency | 55 |
Using only tactics with published evidence behind them.
Your buyers are asking ChatGPT, Perplexity, and Gemini the questions they used to type into Google. The engines answer with a synthesis and a handful of citations. If your site isn’t one of them, a competitor is, and there’s no page two.
Traditional SEO tools can’t see this channel. Ranking #1 on Google and being invisible in AI answers is now a common — and measurable — condition.
A captured engine answer renders here — dated, attributed to its model, citations marked. We don’t mock engine output, so this panel stays empty until we publish a real capture.
ChatGPT
Perplexity
Gemini
Claude
Yes — and some are decent. They’ll tell you your score, show you a leaderboard, and hand you a to-do list. Then you’re alone with the list, or upsold into a $2,000+/month retainer.
Citebook is what comes after the free audit. We deliver the fixes themselves — sections rewritten with real statistics and named sources, schema built from your actual content, every claim verified before it reaches you.
A free audit tells you what’s wrong. Citebook ships what’s right.
| Retrievability | 100 |
| Claim density | 64 |
| Structure | 41 |
| Freshness | 0 |
| Schema validity | 0 |
| Entity consistency | 55 |
No published or updated date is exposed, so engines cannot tell whether this page is current.
No H2/H3 is phrased as a question, so no section maps onto a query an engine is answering.
Body runs 0.65 sourced claims per 500 words against a target of 6.
Top 10 fixes, ranked by evidence strength × effort × page priority. Not everything that could be done — the shortest path to citability, in order. The first three arrive fully written: sourced statistics with publisher, date, and URL attached; question-phrased headings with direct answers; schema generated from your real content, never invented facts.
A shareable web report and the same report as a typeset PDF, emailed to you. Then a 30-minute walkthrough call: we go through it together, and you leave knowing what to do in what order. Everything is yours to keep and implement with your own team.
Because we publish them. Every Citebook score uses these fixed weights, each tied to the strength of the evidence behind it. They live in the codebase as constants — the score cannot drift from the published methodology.
| What we score | Weight | Evidence |
|---|---|---|
| Retrievability Bot access, server-side rendering, status codes. A gate: nothing else matters if engines cannot read the site. | 20 | ● tier A |
| Claim density Statistics, attributed quotes, and externally sourced claims per 500 words. Highest-weight component. | 30 | ● tier A |
| Structure Question-phrased headings, answer-first passages, and self-contained extractable chunks. | 20 | ● tier A |
| Freshness Visible published/updated dates, recency, and absence of stale in-body references. | 15 | ● tier A |
| Schema validity Presence and validity of JSON-LD. Correlational evidence only. | 10 | ○ tier B |
| Entity consistency Brand-name consistency, Organization schema, and named author/founder entities. | 5 | ○ tier B |
Retrievability is a hard gate: if AI crawlers cannot read the site, the total score is capped at 35 and the finding leads the report.
llms.txt “optimisation”
The largest available study found zero correlation with citations, and Google’s own guidance says you don’t need it. We’ll add the file for free if you want it — but we will not charge you for it, and anyone who does is selling you a placebo.
Schema as a magic ranking signal
It’s hygiene with mixed evidence, weighted accordingly — not “the #1 AI signal,” whatever other audits claim.
AI-rewritten everything
Wholesale AI rewrites without sourcing make content less citable. Every fix we ship preserves your voice and carries verifiable sources, or it doesn’t ship.
The first case study publishes when its data is in. We won’t show you numbers we haven’t earned.
If the audit doesn’t show you something you didn’t know about your AI visibility, reply to the delivery email and we’ll refund it.
Free audits measure and recommend. Citebook measures, recommends, and delivers — three fixes arrive fully written with source-verified statistics, ready to paste. We also publish our scoring weights and the studies behind them, which no other audit does.
Those are monitoring platforms — dashboards that track your visibility over time, at $29–$499+/month. They tell you where you stand. Citebook starts with what they end with: the fixing. (Our Monitor tier will handle tracking too, at mid-market pricing.)
Every statistic or quote we inject into your content carries its source URL, publisher, and publication date — and we confirm the source actually contains the claim before the fix reaches you. Anything unverifiable is flagged, never shipped. AI tools are known for inventing numbers; every number we ship is checked against its source first.
No — and be suspicious of anyone who does. AI answers are probabilistic and engines change. What we guarantee: your site will align with every factor the published research links to citation, and you’ll know exactly what changed and why.
Your domain, a paragraph about your business, and your three closest competitors. We handle the rest. No site access required — fixes arrive as copy-paste content and schema.
Because AI engines are writing the answers about every market right now, and most African brands are absent from their own category questions. Early movers get to define the citations everyone else inherits. (Citebook is built in Lagos; we price in USD and work with companies anywhere.)