Visibility in ChatGPT, Gemini and Perplexity answers

Your buyer asked ChatGPT. It named three companies. Yours wasn’t one of them.

We’ll show you which questions leave you out, who gets named instead, and why them. Free, no call needed.

Check your site →Pricing
3×
every question
up to 60% of sources change between runs
19
AI crawlers
checked for access to your site
30–50
questions
in the in-depth audit
$290
audit
credited to your first month
measured · 3 runs ✓

Check your site

9 questions from your category · each asked 3 times · 19 AI crawler signatures

One report per site. We contact only you, and only about this request.

Why it matters

Half your buyers no longer see your search results

A buyer used to type a query and pick from ten links. Today they ask an assistant and get a finished answer: two or three names and a verdict. Only a minority opens the links underneath.

2–3

An answer instead of links

The assistant names two or three companies. Your search ranking no longer decides this: the model builds its answer from sources it trusts.

0.737

Off-site footprint decides

YouTube mentions correlate with getting into answers at 0.737, unlinked brand mentions at 0.664, ordinary backlinks at 0.218 (Ahrefs, 75,000 brands).

+40%

Text built for extraction

Numbers, quotes and source links raise inclusion in answers by 30–40% (Princeton and IIT Delhi, KDD’24).

19

Can crawlers see you

We check 19 AI crawler signatures. If your site is closed to them, the rest of the work counts for nothing.

Classic SEO

  • A fight for the click in a list of links
  • Search ranking as the only metric
  • A report from a single query — guesswork
  • Growth without a before / after measurement

ContentRun · AEO

  • A mention inside the assistant’s answer
  • Share of Answer on a fixed question set
  • Every question asked three times
  • The same measurement every month — a visible difference

What you get

Not a score, but a list of places where you get lost

The first findings are open in full: with the question, the names and the date. You can check each one yourself in a minute.

Example · measured September 5, 2026 · 27 runs
The main question in your category — you’re not named“Which service should I choose for your use case”: 0 mentions in 3 runs. Three competitors are named. The answer is built from four sites — yours isn’t one of them.
The pricing question sends buyers elsewhere“How much does your service cost”: 0 of 3. Two of the named companies have their price taken straight from their pricing page. Your site has no prices, so there’s nothing to quote.
Five of nineteen assistants can’t read your siteTheir requests are refused. Everything you’ve written doesn’t exist for them.
63 more findings: technical 28 · content 19 · off-site footprint 11 · crawler access 5Open the other 63We go through each one in the in-depth audit

SEO, AEO, GEO

Three words — one job

SEO

Get into the list of links

A fight for the click.

AEO

Get into the answer itself

A fight for the mention — there may be no click at all.

GEO

Write so the model quotes you

Numbers, quotes and sources in the text.

We sell it as one job, not three services. Splitting them is marketing, not method.

Pricing

What it costs

Check

Measurement on 9 questions, the two key numbers, the first findings in full, the rest as a count.

free

First month of work

Off-site footprint, page rewrites, publications on the measured topics, technical fixes. At the end — the same measurement and a before / after report.

$2,990

After that

The same every month, with the same questions. Otherwise there’s nothing to compare.

$2,990/mo
The $290 is credited toward your first month. Take the audit and decide to work with us — the month costs $2,700. You don’t pay twice for the same thing.

How we measure

Why you can trust our numbers

  • Every question is asked three times40–60% of sources change from run to run. A report based on a single query is guesswork with nice formatting.
  • The question set is fixedChange the set and any “growth” becomes fake. We version it like code.
  • What we didn’t measure isn’t zeroSuch a line is left out of the calculation and named separately in the report.
  • A proxy measurement is called a proxy“We rank at the top of search” and “the assistant names us” are different claims. We won’t pass one off as the other.

Honestly

What we don’t promise

No share of answers, no timeline for getting in, no growth in AI traffic. The first is unpredictable, the second isn’t in our control, and the third barely moves even when the work succeeds: referrals from assistants are around 1% of visits.

What grows isn’t traffic. It’s the chance you get named at the moment of choice. We promise the work and the measurement: the same questions, the same numbers, a visible difference.

Industries

Industry pages

The same measurement for six industries: where assistants actually take their answers from for real buyer questions, with the source sites named.

Freight & Logistics → Dental Clinics → Online Stores → IT Outsourcing → Law Firms → Construction Companies →

FAQ

Frequently asked questions

Can’t you just “optimize for ChatGPT”?

No, because the answer isn’t built from your site. The strongest signals are external: video, brand mentions, reviews and directories. A plan that is all about your site misses the target.

Who handles my request?

A specialist on our team. Every request is reviewed by a person, and the report is sent to the channel you choose.

My market isn’t English-speaking.

The questions are written in your market’s language, and the measurement runs in it too.

Why don’t you guarantee results?

Because a guarantee comes from someone who doesn’t understand how a model builds its answer. We are accountable for the scope of work and the honesty of the measurement — and show you both every month.

Shall we check whether you get named?

The first measurement is free: 9 questions from your category, three runs each. What happens next is your call.

Check your site →Message us on Telegram