How do I check whether ChatGPT recommends my company?

Ask it directly: "What's the best [your category] tool?" without naming yourself. That tells you what happened once. Because answer engines are non-deterministic — up to 60% of the sources cited change between two identical questions asked the same day — a single check is a sample, not a measurement. A reliable answer needs the same question asked several times across several engines.

Why does an AI engine name a competitor instead of my company?

Usually because it cannot resolve who you are. Before an engine can cite you it has to agree you exist as one specific entity. Most companies that look absent from AI answers are actually being confused with something else that shares their name — a different problem from having too little content, and one that more pages will not fix.

Can any tool guarantee that AI will cite my site?

No. Nobody controls what an answer engine retrieves. What can be done is measure your current state honestly, ship the fixes that make you resolvable and quotable, and re-measure afterwards — reporting the result either way, including when nothing changed.

If someone asked an AI about your category this week, nothing you own recorded it.

No rank moved. No click was logged. Nothing you currently measure will ever tell you about it. CitePack asks the four AI engines whether they name you — and tells you what to do about the answer.

Find out where you stand Or see it for yourself first ↓

Free · No signup · Usually ready in under a minute

First — Don’t take our word for it

Go ask an AI what the best tool in your category is

Whichever one you have open — ChatGPT, Perplexity, Claude, Gemini. Ten seconds. Use your own category, and do not mention your company name.

What’s the best [your category] tool for a [your customer]? Give me your top three.

Most likelyCompetitors, and not you. The buyer who ran that same question this morning saw the same three names. Nothing about it reached your analytics.

Another outcomeSomething that is not your company at all. A different business with your name, a product, an old article. On our first reading, every engine we asked did this to us, and it is a completely different problem from being absent.

If you winIt names you — once, on one engine. That is the weakest kind of good news. The other three may disagree, and the same engine may answer differently in an hour.

Whatever came back, you just found out something you had no way of knowing this morning.

And now you have four questions you cannot answer

01Why did it answer that way? Absent, confused with something else, or simply not the strongest source — identical in a chat window, opposite fixes. And nothing in your analytics can tell them apart.

02Do the other engines agree? You asked one. Your buyers are spread across four, and they routinely disagree about the same company.

03Was that answer stable, or luck? Answer engines are not databases. Ask the identical question twice on the same day and up to 60% of the sources come back different. One reading cannot tell you which you got.

04What would actually change it? The honest answer depends entirely on the first three — and getting it wrong costs a quarter of work that moves nothing.

You just ran one sample, on one engine, once. The free scan reads your page, scores the seven on-page signals, and asks Perplexity one question about your category, then tells you whether your domain was among the sources it cited. The twelve-probe reading, four engines asked three times each, is part of the $49 report. Start with the free one and see whether there is anything here worth paying to look at.

Get the real answer Free · No signup

Second — What CitePack actually is

A scan is a photograph. This is a loop.

Anyone can tell you what an engine said today. That is the easy half, and it goes stale within a week. The part that changes your position is what happens next — and whether anybody goes back to check it worked.

Measure

Free scan: we read your page, score the seven on-page signals, and ask Perplexity one question about your category. With the report: four engines, three samples each, three kinds of question, every individual answer stored.

Free + $49

Diagnose

Which rung you are on and why — absent, confused, resolved but unquoted — plus the rivals named in your place.

Prioritise the fix

The specific thing that rung is blocked by, as deploy-ready pages and identity — not a list of recommendations to interpret.

Re-check

Engines re-index on their own schedule, so we go back weeks later and measure again — and report the result either way, including nothing.

Keep asking

Citations move on their own. Pro keeps asking the same questions and tells you when a result changes, so you hear it from us rather than finding out later. If we could not measure, we tell you that too.

Pro

Steps one and two are free

Enough to know whether you have a problem worth spending anything on — and it is the same measurement we run on ourselves.

Start with step one

Third — The pattern underneath all of it

Getting cited is not a score. It is a ladder, and every rung is blocked by something different

We found this by running it on ourselves first. We had a product, we wanted to know whether the engines knew we existed, and the first real result was not the one we were braced for: we were not invisible — we were somebody else. Every engine answered confidently about a different company that happened to share our name.

We had been about to do what everyone does when they think the engines have not noticed them: publish more. It would have changed nothing. Since then we have seen the same shape in everyone else’s data too — and it explains why so much AI-visibility effort produces nothing: publishing more pages only helps on one of these rungs.

Invisible

The engine has never encountered your company. Ask about you by name and it says so — or invents something.

Where we started

Confused

It reaches something else that shares your name. This is where most companies actually are, and from the outside it is indistinguishable from being invisible — which is exactly what makes it expensive.

Resolved

The engine agrees you exist as one specific company and can describe you. It still does not treat your site as the source.

CitePack is here

Recognized

Ask about you by name and the answer is built from your own pages. Your story gets told in your words instead of a reviewer's.

Nobody yet — including us

You are on one of those five right now

The rung decides what is worth doing next. The free scan tells you what shape your page is in and asks one engine about your category. Your rung comes from the four-engine reading in the report.

Run the free scan

Knowing your rung is worth more than knowing your score. A score tells you how you are doing. A rung tells you what to do next — and, just as valuable, what would waste the quarter. It is also what decides which fix the loop ships you.

3 of 4

engines now identify and cite us by name

0 of 12

category probes name us without being asked

We are on rung four, not rung five, and we publish both numbers at the same size, because a tool that only ever shows you wins is not measuring anything. See every number and how to check it →

Why today rather than next quarter

We do not know how fast these answers harden, and we have not seen a number for it we would stand behind. What we do know is that a reading you have not taken is one you cannot act on, and the first one costs nothing.

So which rung is your company on?

Paste your URL. The free scan tells you what shape your page is in, which of the seven signals need work, and whether Perplexity cited you when we asked it one question about your category. The four-engine reading, your rung, and the pages come with the $49 report.

Find your rung See our own results first →