Service · 01
AI Visibility Diagnostic
$1,000 until Edition 1 of the Index (January 2027), then $1,500 · one-time, prepaid · credited in full against a 60-Day Pilot signed within 60 days · not credited against retainers
Custom and repeated: 40 prompts built with you, three repeats through the official APIs of OpenAI, Gemini, Perplexity and Claude — 480 logged API runs — plus a licensed Google AI Overviews check on your own prompts, reported in its own column; share of voice against five named competitors with variance; the audits; a 90-day plan; and the run data (derived data always, answer text where the engine's terms permit).
- Who it is for
- Operators, brokers, FBOs, MROs, management and sales firms, fuel suppliers and vendors who want to know exactly where they stand before spending on anything.
- What you get
- A panel of 40 buyer questions built with you for your segment, routes, airports or capabilities — custom, unlike the Snapshot's public-panel prompts.
- Three logged runs per prompt per engine through the official APIs of OpenAI, Gemini, Perplexity and Claude (480 answers), each with timestamp, model, measurement surface and answer text; plus a licensed Google AI Overviews check on the same prompts, reported in its own column; the vendor is named on the methodology page.
- Share of voice for you and five named competitors, per engine and per prompt cluster, with variance.
- An entity accuracy audit: how each engine describes your company, and what it gets wrong about fleet, bases, ratings or certificate.
- A crawler, rendering and robots audit of your site, and a page-level extractability check.
- A third-party presence map: the lists, directories, YouTube, Reddit and trade-press pages engines cited, and which of them you are absent from.
- A 12–15-page report, a 45-minute readout with the people who did the work, and a prioritized 90-day plan.
- What you will not get
- A ranking. We report share of voice with variance, because a rank from one run is noise. 1
- Timeline
- Two weeks from panel sign-off.
Measured through official APIs OpenAI API + web search · Gemini API + Search grounding · Perplexity Sonar API · Claude API + web search — the models' retrieval-augmented APIs, not the consumer products. How each surface is measured
Questions about the Diagnostic
Why 40 prompts and three repeats?
Forty prompts cover the buying stages and question types for one segment without padding. Three repeats per API engine is the floor at which a "named / not named" pattern is distinguishable from run-to-run noise; retainers and the Index use five. Google AI Overviews has no answer API, so it is checked through a licensed tool and shown in its own column; Copilot and the ChatGPT app are not measured in the Diagnostic.
Why is the price changing in January?
$1,000 is the introductory price until the first public edition of the Index exists. From Edition 1 the Diagnostic is $1,500. Either way it is credited in full against a Pilot.
Can we choose the competitors?
Yes, five. We will suggest alternatives if the ones you choose never appear in any answer, because that tells you less.
Do we get the raw data?
Yes, as CSV and a private page: extracted names, cited sources and timestamps always; answer text where the provider's terms permit redistribution.