Brand audit across LLMs
A defined prompt set run across ChatGPT, Perplexity, Claude and Gemini. Mention rate, citation rate, sentiment, source-of-citation, all logged.
03 · AEO
Make AI search engines cite you correctly.
Audit how ChatGPT, Perplexity, Claude and Gemini describe your brand today. Trace inaccuracies to their source pages. Fix the sources. Earn the citations. Then measure the picture move on a 30-day window.
The problem
You cannot fix what you cannot see.
A meaningful share of your buyer research now happens inside AI search. Buyers ask ChatGPT to compare your category, ask Perplexity to summarize your category leaders, ask Claude to recommend a fit. The answers shape pipeline whether you measure them or not.
Most B2B SaaS brands have never opened any of these tools in a buyer's voice. They have not seen how often their brand comes up, what gets said, which competitors are cited instead, or which pages the models pull from. So they cannot fix the picture, because the picture is invisible to them.
AI search visibility starts by making the picture visible. We run a defined prompt set across every major answer engine, log mention rate, citation rate and sentiment, and trace each citation back to its specific source page. The audit alone usually reveals three or four corrections worth shipping inside a week.
From there the work is execution: surgical updates to the source pages, schema and citation signals that earn confidence, AEO playbook applied across the relevant clusters, and monthly re-measurement so the work proves itself in numbers. Done well, this earns citation share that traditional SEO cannot reach.
Symptoms we see
- ChatGPT recommends competitors when buyers ask in your category.
- Perplexity describes you using a positioning you abandoned two years ago.
- Claude cites a Reddit thread instead of your own documentation.
- You rank number one in Google for a query that LLMs cite someone else for.
- Nobody knows what your citation share looks like this month versus last.
What's included
Audit, execution, and monthly proof.
Six deliverables, structured as a one-off audit plus an ongoing monitoring loop. Most engagements move into the loop after the audit week.
Source-trace map
Every citation traced back to its source page. The map shows where each LLM is pulling from, which inaccuracies map to which sources, and which pages need surgical updates.
Correction playbook
Page-by-page updates to fix the source. Surgical content edits, schema additions, citation signals, structural changes that move what the model says about you.
AEO playbook for content
The patterns that earn citation across the clusters that matter most. Applied as a structural standard for all new content production going forward.
Monitoring dashboard
A dashboard your team reads in five minutes. Mention rate, citation rate, sentiment, source diversity, all tracked month over month with movement notes.
Monthly re-measurement
The same prompt set re-run monthly. New inaccuracies surfaced. Movement logged against the work shipped. Next month's priorities locked in.
How it runs
Two-week audit, then a monthly loop.
Engagement starts with a fixed-scope audit and rolls into a monthly monitoring retainer. The audit week proves the picture; the loop is where the picture actually moves.
01
Prompt set design
The defined prompts that represent how your buyers actually search inside AI. Pulled from sales transcripts, Reddit threads, G2 reviews and existing search behavior. Locked before measurement starts.
02
Baseline measurement
Run the prompt set across every major answer engine. Log mention rate, citation rate, sentiment, source-of-citation. This is the picture today.
03
Source trace & prioritization
Trace every citation back to its source page. Identify the sources doing damage, the sources doing good, and the gaps where the model has nothing to cite. Prioritize the corrections by impact.
04
Execution
Surgical updates to the source pages. Schema added. Citation signals built. Structural patterns applied. New content commissioned where the gap requires it.
05
Monthly re-measurement & report
The same prompt set re-run. Movement logged. New issues surfaced. Next month's correction list locked.
Who it's for
Built for brands that take AI search seriously.
- Your buyers research in AI tools and your sales team has noticed.
- You want to measure AI visibility properly, not eyeball ChatGPT once a quarter.
- You are willing to ship corrections to source pages when the trace points there.
- You understand AEO is its own discipline, not a Google SEO add-on.
- You want movement you can prove in numbers, month over month.
Pairs well with
What sits around it.
AI search visibility is most powerful with strategy upstream and production downstream. Strategy decides what to be cited for. Production ships the source pages that earn citation.
Frequently asked
Questions we hear.
What is AI search visibility? +
AI search visibility is how often, how accurately and how favorably AI search engines describe and cite your brand when buyers ask questions in their category. It is the AEO discipline measured: prompt-level mention rate, citation rate, sentiment, and source-of-citation, across ChatGPT, Perplexity, Claude and Gemini.
How do you measure AI citation share? +
We run a defined prompt set monthly across every major answer engine, log mention rate, citation rate and sentiment, and trace each citation back to its source page. The output is a dashboard your team can read in five minutes and a movement log that proves the work is shifting the picture.
What if an LLM is describing my brand incorrectly? +
We trace the inaccuracy to its source, which is almost always a specific page or set of pages the model is pulling from. Then we correct the source, add the structure and citations that earn confidence, and re-measure on a 30-day window to confirm the picture has moved.
Is this the same as traditional SEO? +
No. Traditional SEO targets ranking in Google. AI search visibility targets citation in LLM-generated answers. The signals overlap, but a page ranking number one in Google is not automatically the page ChatGPT cites. AEO work requires its own audit, its own measurement, and its own playbook.
What tools do you use? +
Mostly manual right now, by design. The defined prompt set is run by hand across ChatGPT, Perplexity, Claude and Gemini and logged in a tracker. Ahrefs Brand Radar surfaces citation share and AI mention trends. GSC provides adjacent search behavior. In-house Claude agents handle source tracing and analysis at scale. AEO tracking automation is still maturing, so the discipline of a defined prompt set and a baseline is what produces signal, not any tool subscription.
Next step
Let's see what AI says about you today.
A 30-minute discovery call, no pitch, no pressure. We look at what ChatGPT, Perplexity and Claude say about you right now, and whether an AI search visibility engagement makes sense for the quarter ahead.
Book a discovery call →Booking goes straight to Gemma's calendar.