AnswerShelf

Do the machines recommend you?

A ranked read of how often ChatGPT, Claude, Gemini, and Perplexity recommend your brand for buying-intent questions in your category — and exactly which pages earn the mention. Aggregates with sample counts, never a single-shot claim.

$500/ audit · delivered within days
What we check

Six signals, every one grounded.

Small samples render as “insufficient” rather than a shaky percentage — statistical honesty applied to stochastic model output.

Share of voice
What share of high-intent buying questions in your category actually mention your brand.
First-mention rate
Being mentioned isn’t enough — is your brand the first name a model reaches for, or an also-ran.
Recommendation sentiment
When a model mentions you, does it recommend warmly or hedge toward a competitor.
Cited-source domains
The exact pages models cite when they recommend you — the concrete fix list, not generic SEO advice.
Competitor delta
Your share of voice next to your named competitors’, not in isolation.
Week-over-week trend
A longitudinal history a new entrant can’t backfill — the moat is the trendline, not one snapshot.
Sample finding

Not a checklist. An answer.

Illustrative — the format every real finding follows.

critical · share of voice

You appear in 12% of high-intent answers in your category; Competitor X appears in 41%.

Next step: Here are the 6 pages the models cite when they recommend Competitor X — strengthen the equivalent pages on your own site first.

29-point gap to the category leader
Request your audit

Tell us where to send it.

We’ll build your prompt set and run the first battery within days.

We’ll build your prompt set, run the battery across models, and send a report — no account access needed.