You're not in the answer
Buyers ask “best [your category] in [your market]”. The engine cites competitors; you're absent. The deal is lost before your website is ever visited.
AI Visibility Audit · AEO
Your customers ask AI. What does it answer?
Buyers now ask ChatGPT, Perplexity and Gemini — and read one answer that names a few brands and ignores the rest. We measure what the engines say about you, show why, and hand you the fix plan. With evidence for every claim.
best dental clinic for implants in Lisbon
For dental implants in Lisbon, the most frequently recommended clinics areCompetitor A and Competitor B. Both are praised for experienced implantology teams and transparent pricing…
competitor-a.pttop10clinics.comreddit.com
Your clinic — not mentioned. The buyer never sees you.
Invisibility is silent: there is no “we didn't appear in the AI answer” metric in Google Analytics. We make it measurable.
−25%
traditional search volume by 2026, as buyers shift to AI assistants
Gartner
900M
ChatGPT weekly active users — up 125% year over year
OpenAI · 2026
~25%
of Google results pages now open with an AI Overview
Conductor · 2026
−61%
organic click-through on queries where an AI Overview answers first
Seer · 2025
The problem
Three ways you lose deals inside AI answers
Each one is silent, and each one is measurable. Our audit probes for all three.
AI describes you wrong
Outdated services, the wrong category, hallucinated claims. When a machine confidently tells buyers you're something you're not, that's brand damage you can't see.
Buyers arrive pre-decided
The AI recommended someone — so the buyer contacts them, not you. Your pipeline shrinks and sales never learns why.
None of this shows up in your dashboards. There is no “we didn't appear in the AI answer” metric in Google Analytics — which is exactly why the loss goes unnoticed until the pipeline thins.
A different discipline
This is not SEO in new packaging
SEO tools measure the results page. The AI answer sits above it — produced by a different mechanism, on a different axis.
What SEO tooling sees
- Rankings, keywords, backlinks — the layer of links
- AI “mention trackers” built from keyword data — sampled averages, not measurements
- No answer to “what does this engine say on this exact question — and why”
What we measure
- The answer itself — your buyers' real questions, per engine, recorded with cited sources
- Two streams: live retrieval (today's buyers) and model memory (what the AI believes)
- The mechanism behind every gap — so the fix is specific, not a guess
“A page can rank #1 on Google and never be cited by ChatGPT.”
The method
Measured directly. Diagnosed mechanically. Re-measured on a cycle.
Everything runs async — findings, recordings and reports land in your chat. You never book a meeting.
We ask the engines
Real buyer questions across ChatGPT, Perplexity, Gemini, Claude and AI Overviews — multiple runs, neutral accounts, two streams. Every probe stored: prompt, answer, citations, date.
We find the broken layer
Every gap sits on one of four layers. A break high up voids everything below — so we fix from the lowest broken layer, not by guesswork.
You get the plan
A 30/60/90-day roadmap ranked by impact vs cost. Vendor-neutral: hand it to your own dev or content team — or to us.
We re-measure
Same questions, same engines, 30/60/90 days later. It moved or it didn't — you see the numbers either way. A system, not a snapshot.
1AccessCan AI read your site?2StructureCan AI parse it?3SemanticsDoes AI understand you?4AuthorityDoes AI cite you?
The deliverable
See exactly what you'd receive
No generic AI slides. A measurement report your team can act on the same week:
- What each of the 5 engines actually says — with the sources it cited
- Competitor citation map: who gets named instead of you, per engine
- 4-layer diagnosis with the root cause, not a list of symptoms
- 30/60/90-day roadmap ranked by impact vs cost, with a re-measure date
Where we draw the line
“We can't force AI to cite you.
We remove the reasons it doesn't.”
We don't promisethat an LLM starts citing you tomorrow, a #1 spot in any engine, or leads and sales. Nobody honestly can — the engines aren't ours to control.
We do promisemeasurement you can verify, fixes you can inspect (schema diffs, validators, crawl checks), and a plan specific enough for any team to execute.
We re-measurethe same questions at 30, 60 and 90 days and show you what moved. If a vendor guarantees rankings inside AI engines, that's your cue to leave.
Pricing
Start free. Every step fixed-scope.
Diagnosis first, always — without it, fixing is guesswork. Everything is delivered async to your chat.
$0
See what AI says about you — the killer number, the worst quote, your gap page.
- 5 buyer questions × 1 engine
- Screen recordings of the answers
- Personal gap page link
$250
One engine, in depth — including what it believes about you from memory.
- ~15 buyer questions, dual-stream
- Hallucination check
- Top-3 fixes, ranked
$500 intro
list $990
The complete measurement: every engine, every layer, and the plan.
- 5 engines × 20 buyer questions
- 4-layer diagnosis + competitor map
- 30/60/90 roadmap, impact vs cost
- Evidence archive + re-measure gates
$1,000+
We implement what the audit found, then keep measuring so gains don't rot.
- Foundation · Shield · Engine packages
- Schema, rendering, entity, authority
- Monthly re-measurement retainers
Fixed scope · fixed price · fixed SLA. No re-scoping mid-flight, no retainer required to buy an audit.
Questions
Fair questions, straight answers
What is an AI Visibility Audit, and how is it different from SEO?
SEO optimizes your position on a page of links. We measure the answer above that page — what ChatGPT, Perplexity, Gemini, Claude and Google AI Overviews actually say and cite when buyers ask about your category. Rankings and citations are different mechanisms: a page can rank #1 on Google and never be cited by an AI engine. The industry calls this AEO or GEO; we call it measurement.
Which AI engines do you cover?
ChatGPT, Perplexity, Gemini, Claude and Google AI Overviews — each in two streams: live retrieval (what buyers see today) and model memory (what the model believes from training, measured via API with web access off). The engine list is kept current as the market moves.
AI answers vary. How can you measure something random?
Single answers vary — that's why every question runs multiple times per engine on neutral accounts, and we report frequencies, not one-offs. “You appear in 0 of 116 runs” is not noise. Every probe is stored — prompt, full answer, cited sources, date — so you can check any claim in the report against the raw evidence.
Can you guarantee that ChatGPT recommends us?
No — and you should distrust anyone who says yes. Nobody controls these engines. What we control: every technical and semantic reason AI currently skips you. We remove those reasons, then re-measure the same questions at 30, 60 and 90 days and show you exactly what moved.
Can we implement the fixes ourselves?
Yes. The roadmap is vendor-neutral by design: every fix is specific enough to hand to your own developer, content team or agency. Many clients implement in-house and come back only for the re-measurement. If you'd rather we do it, fixed-scope packages exist for exactly that.
How fast do results show?
Three clocks. Mechanical fixes verify immediately — schema validators and crawl checks pass or they don't. Citation changes in live-retrieval answers typically take 30–90 days. Model-memory changes follow the engines' training cycles — quarters, not weeks. Anyone promising next-week citations is lying to you.
Why don't you do calls?
Because you don't need a meeting to see evidence. Everything runs async in chat — WhatsApp or email. You get artifacts you can forward and re-read: findings, screen recordings of the engines answering, your gap page, the full report. Ask anything in the thread; we answer fast.
Free · 48 hours · no meetings
Find out what AI already tells your buyers
Send your domain and category. Within 48 hours you get the findings, the recordings and your gap page — in your WhatsApp or inbox, with receipts.
Get my free AI snapshot