When buyers ask AI for a shortlist, are you on it?
A growing share of buying research never touches a results page: someone asks an AI engine for the best option and reads one answer. If you are not in that answer, a competitor is, and no analytics tool shows you the deals you never entered. This audit makes the invisible measurable.
What you get
The method
Derive the buyer queries
From your market and site, we build the question set a real buyer asks an engine: brand checks, best-of lists, category and review queries.
Probe the engines, capture everything
Each query runs against each engine and the full answers are captured as evidence, including retrieving the actual sources behind Google's AI Overviews rather than assuming them.
Detect presence deterministically
Cited, mentioned or invisible is decided by matching against the captured answers, not by an AI's opinion of an AI. What cannot be measured is declared a gap, never guessed.
Score by fixed rubric
The four pillar scores compute from the evidence by rules, so the same evidence always produces the same score. Your next audit is comparable to this one.
Verify, then the gate
An automated verifier recomputes every score and checks every claim against evidence, then the adversarial human review. Reports that cannot prove themselves do not ship.
Who this is for, and who it is not
It is for brands whose buyers research before buying: SaaS, services, anything with a comparison phase. If your customers ask ChatGPT for recommendations, and they do, this is the visibility you cannot see from Search Console.
It is not for pre-launch sites with no footprint yet; the engines have nothing to cite and the audit would mostly measure absence. Build the foundation first, the Growth Plan sequences that correctly.
What happens after checkout
Scoped at checkout
Your market, the competitors worth tracking and any queries you wanted included were captured in the order form. Nothing else needed.
Probes run
Engines are queried, sources retrieved, evidence captured. Progress visible in your dashboard.
Verification and QA
The automated verifier, then the hostile human review.
Delivery and your approval
Report in your dashboard, up to three revision rounds included.
What this replaces
| Alternative | Typical cost | What you actually get |
|---|---|---|
| "AI visibility" tool subscriptions | $100 to $400/mo | Volume metrics without judgment or a roadmap |
| Asking ChatGPT about yourself | Free | One engine, one day, no rubric, no competitors, no reproducibility |
| Agency AEO retainers | $2,000+/mo | Often this exact audit, spread across a quarter |
Questions, answered straight
Which engines do you measure?
Claude, Gemini and Google AI Overviews with full source retrieval today. Engines we cannot yet measure are listed as explicit gaps in your report, because guessing would poison the score.
How do you decide if I am "cited"?
Deterministically: captured answers are matched for your brand and domain. No model grades another model. The same evidence always produces the same verdict.
What if the news is bad?
Then you learned it for $149 instead of discovering it in pipeline numbers. The report pairs every gap with the fix, and the strengths section shows what to build on.
How often should this be re-run?
Engines drift monthly. One audit sets the baseline; AI Citation Monitoring at $49/mo watches the drift and flags misrepresentations.
Can you fix what the audit finds?
Yes: AEO Optimization implements the entity, schema, llms.txt and content work, then re-measures against this baseline.