ChatGPT, Gemini and Claude now answer "which provider should I trust?" and "which software should I buy?" with a short list. If you are not on it, that customer never sees you. We find out how often you are named, and why.
Credited to the work if you continue with us within 30 days.
When someone asks an AI who to choose, it names a few businesses. We find out how often it names yours, and what it is reading about you.
Businesses and numbers in the animation are made up for the example.
There is no single score. An AI answer draws on everything it can read about your business. These are the signals we check, grouped by where they come from. The exact list changes with the industry.
Does the AI know what you are?
Reviews and ratings.
Articles, lists and directories.
What real people discuss.
What your own pages say.
Can the AI read you at all?
That is 30 signals in six groups, plus the extras for each industry.
Each AI weighs its sources differently. A business that is named often by one can be missing from another, so we test them side by side.
| Model | Names you | Average position | Cites your site | 30 days |
|---|
Example values for one made-up business.
We look at the review sites, directories and competitor pages that AI reads. The report shows how often you are named, who is named instead, and which pages the AI is reading.
All figures are examples.
We ask your question 50,000 ways across more than 10 AI models and record every answer: who gets named, in what order, and which pages the AI read.
A plain list of what to change: wrong facts, thin reviews, missing articles, pages the AI cannot read. Ordered by what matters most.
Once the changes are made, we run the same questions again and show what moved. Up, down or the same.
We ran our method on one question: "Which broker is best to trade silver?" This is a sample report from our own test run, not a result for a client. The PDF says so on its first page. Click it to open.
The broker names are real. The numbers show what the test model said on the day, not what is true of any company, and they are a test of our method. We will add real client reports here, with permission, once we have them.
Leave your email and we send you a sample report. Want a first look at your own business? Add your business and the question your customers ask, and we send you a one-page snapshot.
Every figure shows its date and how many answers are behind it.
We check 30+ signals to find why you are or are not named.
The full report ships with the data, so your team can check our work.
$750 for the audit, credited to the work if you continue. Fixed quote before you commit.
The full audit shows how often AI names you, who it names instead, and why, with a written fix list and a fixed quote for the work. If you continue with us within 30 days, the $750 comes off the price.
The full measurement across 10+ models, the signals behind each result, a written fix list and a fixed quote for the work. Raw answers included, so your team can check our work.
We work the fix list: wrong facts, reviews, listings, content and technical access. We re-test at set points and show what moved. Priced by scope, with a fixed quote before you commit.
Engagements are paid in instalments tied to what we deliver, never to a ranking. We do not promise or charge for a position.
Not everyone wants a PDF. Here is what one report could look like as data your own team or tools can read. On the left, what you send. On the right, what comes back.
Coming soon: this data service is not live yet. The address and fields are placeholders.
Generative engine optimization. It means working on the things AI assistants read, so that they know your business, describe it correctly and name it when people ask.
They overlap, but they are not the same. SEO is about where your site appears in a list of links on Google. GEO is about whether an AI names your business inside the answer it writes. AI reads the same web pages, plus reviews, forums and directories.
No. Nobody can, and you should be careful with anyone who says they can. AI answers change from day to day and no company controls them. What we do is measure where you stand today, show what the AI is reading about you, fix what is wrong or missing, and test again so you can see what moved. Sometimes the numbers go up. Sometimes they stay flat.
If people ask AI about it, yes. We test the questions your own customers would ask, whether you are a broker, a clinic group, an insurer, a software company or a hotel chain.
A report showing how often AI names your business, who it names instead and which pages it reads. A list of what to fix, in order. And a second test after the changes, so you can compare.
The AI gives a different answer each time, and people ask the same thing in many ways. One answer tells you almost nothing. Tens of thousands of variants give you a rate you can trust.
More than ten, covering the main chat assistants and AI search features. Each one picks its sources differently, so the report shows results for each model.
Mostly things outside the AI itself: wrong or missing facts on other sites, reviews, articles and lists, forum threads, and the content and technical setup of your own website.
It depends. AI systems update on their own schedule, and some changes take weeks to show up. We will not give you a date. We re-run the same questions and show you what changed.
It helps, but it is not required. AI also reads reviews, maps, directories, news and forums, and we can work on those.
The full audit is $750, and we deduct it from the price if you continue with us within 30 days. Larger engagements are priced by scope. Contact us and you get a fixed quote before you commit. See the pricing section above.
Leave your email and see a sample report, or buy the full audit for $750 and we start straight away.
Buy the full audit, $750 See a sample report