Read the guideGet an audit
Open research · The Lab’s benchmark

The Citevane Index

A single, named benchmark for one hard question: how reliably do AI answer engines retrieve, read and cite this site? The Index turns that into a position you can track — where a site sits on the climb from merely crawlable to consistently quoted.

AI Visibility Lab Research Team
Reviewed by the editorial desk
Published June 20266 min read

Why publish a benchmark

“Are we more visible to AI than we were?” is a fair question with no good answer, because the thing being measured — retrieval, readability, citation — is spread across three very different layers. Naming a single benchmark, the Citevane Index, lets us collapse those layers into one position you can point at and watch move. It’s the measuring stick for the same progression the SEO vs GEO experiment is built to test — a benchmark, not a slogan.

A model tells you what to fix; a benchmark tells you whether the fix worked. The Index is the second half of that pair.

What the Index reads

The Index weighs three bands of signal, in the order an answer engine itself works. Each band maps onto a rung of the Crawl-Read-Cite Ladder, and each depends on the band beneath it — which is why a site can’t score high overall on the strength of the top band alone.

  1. Retrieval — can engines reach and fetch the page at all?

    The first band scores reachability: whether a crawler can request the URL and get the page back. A site that’s blocked in robots.txt, marked noindex, orphaned, or slow/erroring drags the Index down here — nothing above this band can count if the engine never fetches the page. This band corresponds to the Crawl rung of the Ladder.

  2. Readability — is the answer in the initial HTML, structured to parse?

    The middle band scores legibility: whether the facts that matter are present in the initialserver HTML — not painted in later by JavaScript — and laid out with real headings and matching structured data. A page whose content only appears after hydration scores low here even if it looks fine in a browser. This band corresponds to the Read rung of the Ladder.

  3. Citation — does the page carry the signals that earn a quote?

    The top band scores quotability and trust: a direct, self-contained answer an engine can lift; a clear author and publisher entity; structured data that matches the visible text; and corroboration from sources the engine already trusts. A site only scores high on the Index overall once this band holds and the two beneath it do. This band corresponds to the Cite rung of the Ladder.

The Index and the Ladder

The Citevane Index and the Crawl-Read-Cite Ladder are a matched pair: the Ladder is the model (three rungs a page climbs to earn an AI citation), and the Index is the benchmark that measures it (how far up those rungs a site has actually climbed). Read the Ladder first to learn the rungs; come back to the Index to track your position on them over time.

Low on the Index

The site stalls in the Retrieval band — blocked, orphaned or noindex’d. The engine rarely fetches it, so nothing above can count.

Mid on the Index

Fetched and read, but rarely quoted: the facts are in the HTML, yet there’s no self-contained answer or trust signal worth citing. Work lands in the Citation band.

High on the Index

Retrieved, read and cited: a direct answer, matching schema, a real publisher entity, outside corroboration. All three bands hold.

The bands map cleanly onto a real audit. Our worked AI visibility audit example runs five checks that sort straight onto the Index: View Source and indexation feed the Retrieval and Readability bands; schema, listing consistency and a plain-text answer feed Citation.

The honest caveat

The Citevane Indexis a benchmark under test, not a certified grade. It’s the measuring stick behind our live two-site experiment, which is still in its baseline window — so we’re publishing the framing now and the findings as the data lands. How heavily each engine weights retrieval, readability and citation is reported rather than published and shifts month to month, so treat the Index as a durable way to track AI visibility rather than a number any single engine hands you.

Frequently asked questions

What is the Citevane Index, and who publishes it?
The Citevane Index is AI Visibility Lab's benchmark score for how reliably AI answer engines retrieve, read, and cite a website. It's published by the AI Visibility Lab Research Team — the editorial and research group behind this site — as a single lens for tracking where a site sits on the Lab's Crawl-Read-Cite Ladder: the higher the Index, the further a site has climbed from merely being crawlable to being consistently quoted in AI answers. It reads three bands of signal — Retrieval, Readability and Citation — that map onto the Ladder's three rungs. It is a research lens for tracking progress over time, not a ranking guarantee or a number any single engine hands out.
How is the Citevane Index different from the Crawl-Read-Cite Ladder?
They're a matched pair. The Crawl-Read-Cite Ladder is the model — three rungs a page climbs to earn an AI citation. The Citevane Index is the benchmark that measures the Ladder: it scores how far up those rungs a given site has actually climbed. The Ladder tells you what to fix and in what order; the Index gives you a single position to track as you fix it.
Is the Citevane Index a score, and can I get a number for my site?
The Citevane Index is a benchmark — a position, not a precise grade. We describe it as a band on the Ladder (closer to crawlable, or closer to consistently cited) rather than a decimal a single engine certifies, because how heavily each AI engine weights retrieval, readability and citation is reported rather than published and shifts over time. The honest use of the Index is to track your own direction of travel across the three bands, not to chase a leaderboard number.
Has the Lab proven the Citevane Index with data?
Not yet, and we won't pretend otherwise. The Index is the benchmark behind our live, two-site experiment, which is still in its baseline window — so we're publishing the framing now and the findings as the data arrives. Treat it as a useful, testable lens for tracking AI visibility, not a proven law. How each engine weights these signals is reported rather than published, and it shifts over time.

Keep going

The Crawl-Read-Cite LadderThe model the Index measures — the three rungs a page climbs, in order, before an AI engine can quote it.SEO vs GEO: the experimentThe live, two-site experiment the Index benchmarks — the hypothesis, the method, and what we’re measuring.AI visibility audit exampleFive concrete checks, worked end to end — each one feeds a band of the Index.
See where you land

Find your position on the Index

A plain-English walkthrough of how to test whether ChatGPT and other answer engines can retrieve, read and cite your site — no account required.

Test your AI-search visibility