When someone asks an AI who to hire, do they hear your name?
Your customers are asking ChatGPT and Gemini for recommendations instead of searching. Those answers name three or four companies. We find out whether you are one of them, who is being named instead, and what to change.
One client's finished check. Twelve answers read per name. Names withheld.
Ask one question now
We put your question to a live assistant and read the answer back.
Three things people get wrong
Asking once proves nothing
Ask the same assistant the same question twice and you can get different names. One ask is a screenshot. We ask every question three times and count who shows up, so the result is a measurement.
Arabic is a different answer
The Arabic answer is built from different websites, and it often names different companies. Half your market asks in Arabic. Every tool we know of tests English only. We test both.
Ranking on Google does not carry over
Assistants build answers from a different set of sites than the ones that rank you on Google. Companies on page one of Google are routinely absent from AI answers, and the other way round.
The same question, in two languages
From a finished check: one company, one question about their exact service, asked twelve times in each language.
Asked in English
Named in 7 of 12 answers.
Asked in Arabicبالعربية
Named in 0 of 12 answers. Their Arabic-speaking customers never hear their name.
Pricing
Fixed prices, fixed turnaround, and a report you can hand to whoever runs your website.
Quick check
$450
Delivered in 3 days
12 questions, asked twice each, in English.
96 answers read.
Most clients choose this
Full check
$1,800
Delivered in 5 days
40 questions, asked three times each, in English and Arabic.
480 answers read.
Check and fix
$6,500
Delivered in 3 weeks
Everything in the full check. We implement the changes, then re-check at day 21.
960 answers read, across two full checks.
Already had a check? A quarterly re-run repeats your full question set every three months and reports what moved, for $2,500 a month. Ask us when your first report is delivered.
Questions we get asked
Which AI assistants do you test?
ChatGPT, Gemini, Perplexity and Claude. We report each one separately, because the answers differ and so do the fixes. We never blend them into a single score.
Why do you ask every question three times?
Because the answers change between asks. A company can be named on the first ask and absent on the second. Asking three times and counting is the difference between an anecdote and a result you can act on.
Can you guarantee we get recommended?
No, and nobody honestly can. The assistants decide what to say. What we can do is show you exactly where you stand, which websites the answers are built from, and the specific changes that put you in the running.
Do you write the questions, or do we?
We draft them from how your real customers phrase things, in English and in natively written Arabic. Then you review every question, edit or replace any of them, and nothing runs until you approve the list.
Is this what people call AEO or GEO?
Yes. Those are the industry's names for this work. We avoid the acronyms because they describe how the work is done rather than what you get. If you want the full picture, we wrote it up plainly here.