Presence
When the buyer’s question doesn’t name you, does the answer include you?
Measured on unbranded prompts.
The proof layer for AI answers
See why AI recommends your competitors.
AnswerMap runs your buyers’ questions through ChatGPT, Gemini, Perplexity and Claude, traces the sources behind every recommendation, and turns the gaps into page-level actions.
Free sample: one prompt-level finding from ChatGPT. Full analysis from US$4,500 (launch pricing).
ChatGPT · sample finding
Buyer prompt · unbranded
Which [category] platforms should a mid-size company shortlist?
Answer shortlist
- 01Competitor Arecommended
- 02Competitor Bmentioned
- —Your brandmissing
Why Competitor A wins
AI claim
“Competitor A is the safer choice for mid-size teams.”
- Third-party comparison listsupports
- Competitor A case studysupports
- Your comparison pagenot cited
- Who appears
- Competitor A, Competitor B
- Who wins
- Competitor A
- The gap
- Your proof isn’t cited
- ~200buyer prompts, agreed at kickoff
- 4engines: ChatGPT, Gemini, Perplexity, Claude
- 3×runs per prompt, per engine
- 90 daysrerun of the same prompts
Why it matters
Buyers ask AI for a shortlist. Your dashboards don’t show the answer.
Answer engines name a few vendors, recommend one, and cite the pages that convinced them. If you’re left out, or named but not recommended, it won’t show up in your rankings or traffic reports.
- Category
Best [category] platforms for a 500-person company?
- Alternatives
What are the alternatives to [incumbent]?
- Comparison
[Vendor A] or [Vendor B] for [use case]?
Each answer decides three things: whether you’re in it, whether you’re the one recommended, and which sources it trusts.
How we measure
Two numbers, then the cause.
AI answers change from run to run, so one screenshot proves little. We run every prompt 3 times in each of 4 engines, report every rate with a 95% confidence range, and label each finding robust or directional. We tell you which findings to bet on.
Preference
When the answer names you, does it recommend you?
Measured on the answers that mention you.
The cause
Which sources does the answer cite, and what does it claim about you?
Cited sources, owned-citation share, and AI claims checked against your core pages.
- Claim matches your core pagessupported
- Claim missing from your pagesmissing
- Claim contradicts your pagesdisputed
- 3 runsper prompt, per engine
- 4 enginesChatGPT, Gemini, Perplexity, Claude
- 95%confidence range on every rate
-
- robust
- Strong enough to act on.
- directional
- A signal to watch or test, not yet a bet.
Anonymized case
A presence problem, not a persuasion problem.
The client: a B2B device-management challenger after a rebrand.
Presence
gap40%
of unbranded answers included the brand (30 of 75). The incumbent’s product suite appeared in 84% (63 of 75).
Preference
strength47%
of its mentions were recommendations (14 of 30), more than any rival. Nearest rival: 33%. The incumbent: 8%.
Old brand name
stale30 of 30
unbranded mentions still carry the old brand name. Old-name URLs were cited 32 times; the new domain, 21.
Owned citations
weak3%
of unbranded citations pointed to the brand’s own pages (21 of 653). Competitor-owned pages took 50%.
The reading: when ChatGPT mentions this brand, it recommends it at a higher rate than any rival. The gap is getting into the answer, and the sources ChatGPT reads still point to the old name.
Free and paid
Start with one free finding. Go deep when it’s worth it.
The free sample shows you one prompt-level finding in your category. The full analysis measures the whole shortlist and hands your team the plan.
AI shortlist sample
We run buyer prompts for your category in ChatGPT and show you one prompt-level finding: who appears, who wins, and the source or content gap behind it.
- Engine
- ChatGPT
- Prompts
- Buyer prompts for your category
- You get
- One prompt-level finding
- It shows
- Who appears, who wins, and the source or content gap behind it
Full AI shortlist analysis
Full analysis from US$4,500 (launch pricing)
The 90-day rerun is included.
- About 200 buyer prompts, agreed with you at kickoff
- ChatGPT, Gemini, Perplexity and Claude: 3 runs per prompt per engine
- 3–5 named competitors
- Presence and recommendation rates with 95% confidence ranges
- Citation sources and owned-citation share
- AI claims checked against your core pages
- Every action as a page-level brief: target prompts, page type, claims to support, sources to earn
- Raw evidence export
- Rerun of the same prompts at 90 days, with a change report
Public sample
See the format before you ask.
A public sample for the product analytics category: 15 buyer prompts across all four engines, presence vs preference with 95% ranges, cited sources, claims checked, and three page-level briefs.
About AnswerMap
We measure and plan. We don’t sell content hours.
AnswerMap is an independent service that measures how AI answer engines treat your brand and turns the gaps into a plan your team can ship. Because we don’t sell content hours, nothing in the plan is there to create work for us. Your team, or the agency you already use, does the work; we rerun the same prompts at 90 days to show what moved.
- 01
Agree the prompts
About 200 buyer prompts and 3–5 named competitors, agreed with you at kickoff.
- 02
Run and measure
3 runs per prompt in ChatGPT, Gemini, Perplexity and Claude. Presence and preference, each with a 95% confidence range.
- 03
Trace the cause
Cited sources, owned-citation share, and AI claims checked against your core pages.
- 04
Write the briefs
Every action as a page-level brief, in priority order, with the raw evidence export.
- 05
Rerun at 90 days
The same prompts again, with a change report that shows what moved.
- Target prompts
- Alternatives to [incumbent]; best [category] for [segment]
- Page type
- Comparison page
- Claims to support
- [What AI says, or misses, about you]
- Sources to earn
- [Third-party list cited in the answer]
- Check
- Same prompts, rerun at 90 days
- Independent. We measure and plan; we don’t sell content hours.
- Checkable. You get the raw evidence export, not just our summary.
- Honest about uncertainty. Confidence ranges, and every finding labelled robust or directional.
Contact: hello@answermap.io
Free AI shortlist sample
What does AI say about your category?
Request a free sample and we’ll show you one prompt-level finding from ChatGPT: who appears, who wins, and the source or content gap behind it.