ChatGPTLive web searchLive
When buyers ask AI for a supplier, does your name come up?
Robin Knows asks the assistants your buyers ask, the same way every time, then reports who gets named, who gets cited, and where you are missing.
Measured across ChatGPT, Claude and Google AI Overviews
Who are the best suppliers of [your category] for small manufacturers?
Three suppliers come up most often:
- 1Competitor AUsually recommended for volume orders
- 2Competitor BStrong on catalog range
- 3Your brandKnown for fast lead times yourbrand.com/guides
We do the asking. You do the deciding.
Each engine gets the same questions with live web search on, so what you read is what a buyer would have been told.
ClaudeLive web searchLive
Google AI OverviewsSampled monthly, reported on its ownLive
Perplexity and GeminiBuilt, switching on as validatedSoon
Sample questions
Who makes [your product] for outdoor use?Best [category] supplier with fast lead times[Category] vendors comparedAlternatives to [Competitor A]How do I choose a [category] supplier?Is [Your brand] a good [category] supplier?Which [category] brands do engineers trust?Who has the best prices on [your product]?
Robin Knows shows brands what AI says about them, using the same questions every time, and keeping every answer word for word.
Questions written the way buyers ask
About 50 of them, versioned and frozen for the quarter. Changing the questions would break the trend.
Three runs, averaged
One answer is an anecdote. Three smooth the noise without hiding it.
- Run 1Named
- Run 2Absent
- Run 3Named
- Averaged2 of 3 · sample
Quotes are never edited
What an engine said is evidence, so it is shown exactly as returned.
Provenance on every panel
Tracker, Analytics, Search Console or sample. If it is not measured, it says so.
Four steps. Zero guesswork.
- 01
We write the questions
About 50 prompts, phrased the way your buyers ask. Frozen for the quarter so trends stay comparable.
- 02
We ask each engine
Live web search on, three runs per question, averaged to smooth the randomness in model answers.
- 03
We parse every answer
Brands, order, sentiment and cited sources become structured data. The original text stays attached.
- 04
You read the trend
A dashboard and a monthly readout, with the verbatim answer one click from any number.
Three suppliers come up most often: Competitor A, Competitor B, and Your brand. Competitor A is usually recommended for volume orders…
Read it three ways.
The dashboard
Every panel, every engine, always current.
- Mention rate, share of voice, citation rate
- A provenance tag on every panel
- Filter by engine and by question
The monthly readout
The headline, written for the person who has to explain it.
- Where you stand and what changed
- Comparable periods only, never a partial window
- Verbatim answers one click away
The gap report
The questions you never appear in, and who does.
- Competitors named in your place
- The pages the engines cite instead
- A content plan that falls out of it
Find out what the assistants say about you.
We are onboarding a small group of teams first. Tell us about your brand and we will be in touch about a first report.
Read by a person.
We look at your brand and category, then email you if there is a fit. No payment details are collected.