Nakama Growth's AI visibility audit runs your real buyer prompts through AI engines, repeats each run, and records where your brand is mentioned, recommended or cited next to named rivals. You get dated evidence, the sources engines trust and a fix list. Start with a free sample or book the full audit.
Engines agreed for your audit, each reported separately.
AI visibility sampleSample, redactedPrompt 1 · “best tool for teams”
| Engine | Mentioned | Recommended | Cited |
|---|---|---|---|
| Engine A | 2 of 3 | 1 of 3 | 1 of 3 |
| Engine B | 3 of 3 | 2 of 3 | 0 of 3 |
Example data. Engines reported separately. No blended cross-engine number.
An AI visibility audit is a repeated test of how AI engines answer the questions your buyers ask. For each prompt, it records whether your brand is mentioned, recommended or cited as a source, which competitors appear instead, which sites the engines pull from, and what they get wrong about you.
Mentioned, recommended and cited are counted separately. Never merged. No percentages, scores or grades.
What it shows
Whether your brand is named anywhere in the answer.
What you do with it
Find the buyer prompts where you are absent and start there.
What it shows
Whether the engine suggests you as an option to choose.
What you do with it
Tighten the pages and proof that explain who you are best for.
What it shows
Whether one of your pages is linked as a source.
What you do with it
Make the cited-worthy pages clear, current and easy to quote.
What it shows
Which rivals show up on the same prompts, and how often across runs.
What you do with it
Learn which competitor positions you need to answer.
What it shows
Which third-party sites the engines pull from for your category.
What you do with it
Pick the review, list and media sites worth earning a place on.
What it shows
What engines get wrong about you, and whether their bots can read your site.
What you do with it
Correct the facts at the source and unblock the pages that matter.
We show where you stand today, what is likely driving it, and what to fix.
Engines reported separately. No blended cross-engine number.
Which third-party sites the engines pull from for your category.
A call to walk the findings, a ranked fix list and a date to run it again.
| Engine | Run date | Mentioned | Recommended | Cited |
|---|---|---|---|---|
| Engine A | 14–16 Oct | 2 of 3 runs | 1 of 3 runs | 1 of 3 runs |
| Engine B | 14–16 Oct | 3 of 3 runs | 2 of 3 runs | 0 of 3 runs |
Engines reported separately. No blended cross-engine number.
Lighter stepFree sample | Full measurementPaid AI Visibility Audit | |
|---|---|---|
| Prompts | 1 to 3 prompts (yours + up to 2 we write) | Agreed buyer prompt set |
| Engines | 2 engines, reported separately | Engines agreed for your audit, each reported separately |
| Competitors | Up to 3 you name | Named rivals + new names that recur |
| Runs | 3 runs each, per engine, dated | At least 7 dated runs per prompt per engine |
| Source map | Top gaps only | Full citation source map |
| Average position | Not included | Yes, per engine |
| Accuracy check | Not included | Yes, per engine |
| Roadmap | Not included | Prioritized fixes + re-test date |
| Walkthrough call | Not included | Yes |
| Delivery | Emailed, not instant | Report + call |
Free sample
Scope: 1 to 3 prompts × 2 engines × 3 runs each, per engine. No score, no %.
Emailed, not instant. A person runs your prompts and checks each answer before we send it. No score, no average position.
We agree the buyer prompts and named competitors with you before any runs.
Each prompt runs at least 7 times per engine, with every run dated.
We map mentions, recommendations, citations, sources and errors, engine by engine.
A call to walk the findings, a ranked fix list and a date to run it again.
Pricing on a call
Want us to fix what the audit finds? We run GEO and AEO work after the audit and re-test the same prompts.
No. Nobody controls what AI engines say. We show where you stand today, what is likely driving it, and what to fix. Then we re-test the same prompts so you can see what changed.
AI engines can give different answers to the same prompt from one run to the next. In the paid audit we run each prompt at least 7 times per engine and date every run, so you see a pattern instead of one lucky or unlucky answer.
A mention is your brand named in the answer. A recommendation is the engine suggesting you as an option. A citation is your page linked as a source. We count all three separately because one can happen without the others.
Most graders run a prompt once and return a single score. We repeat runs, report each engine on its own, show the exact answers and sources, and hand you a ranked fix list with a walkthrough call.
Your site, the buyer prompts you care about, and the competitors you want to compare against. If you are unsure about prompts, we draft a set with you before any runs.
You get the report, a walkthrough call and a prioritized roadmap with a re-test date. You can run the fixes yourself or ask Nakama Growth to handle the GEO and AEO work.
The paid audit covers the engines agreed for your audit, each reported separately. The free sample uses 2 engines.
Yes. You get 1 to 3 prompts across 2 engines, with 3 runs each per engine, emailed to you. It is a sample, not a full measurement, so it has no score and no average position.
It depends on how many prompts and engines are in scope. We confirm the timeline when we agree the prompt set.
Pricing depends on the number of prompts, engines and competitors in scope. Book a call and we will quote the audit, or the audit plus fixes, for your site.
Book the full audit, or start with a free sample emailed to you.