Yes. There are two, and the free one you run yourself is better than most people expect, as long as you avoid the mistake almost everybody makes with it.
The mistake is asking once.
Method one: ask it yourself, properly
This costs nothing and takes about fifteen minutes. Open a private window and log out of everything first, because an assistant that knows who you are is not showing you what a stranger sees.
Then ask five questions the way a couple would type them, not the way you would describe yourself.
- Recommend a wedding photographer in your city, using your actual city.
- What does a wedding photographer cost in your city, at the size of wedding you usually shoot.
- A question naming a venue you have worked at, asking who shoots there.
- A question about your actual style, in plain words rather than your tagline.
- The question a couple asks when they are ready to book, whatever that is in your category.
Swap photographer for florist, planner, venue or DJ as needed. Ask each one in at least two services, ChatGPT with web search on and Perplexity are the easiest pair to start with, and ask each question three times.
Why three times, in numbers
In September 2026 I measured a wedding florist in Stockholm on sixteen questions, three runs each, on four services. One hundred and ninety two answers. She appeared eight times. Every single one of those eight was a single run out of three, and not one repeated.
Ask each of those questions once and roughly half of her appearances would have been reported as wins and half as losses, decided by nothing but which run you happened to catch.
A venue measured the same month did the same thing on a larger scale: twenty three appearances across seventy two question and service pairs, and not one of them stable across runs. Across twenty measured photographers there were eighty nine of these near misses.
8 of 8
florist appearances that came in exactly one run out of three
23
venue appearances, none of which repeated across runs
89
near misses across twenty measured photographers
What to write down
Keep it boring and consistent, because the value is in comparing it to next month. For every question and every service, note three things.
Were you named, and in how many of the three runs. Two or three out of three is a real position. One out of three is a near miss, which is the most useful category of all because it is where small changes show up first. Zero is zero.
Who was named instead. The same four or five names will keep coming back. That list is your actual competition in AI answers, and it is often not the list you would have written yourself.
What sources it cited. Where the service shows them. This is the part that tells you what to do next, because it shows which pages the answer was assembled from.
The catch with doing it manually
Two of the four services make this hard. Gemini returns no readable source list at all, only redirect links, so you can see that it named somebody but not why. And doing five questions times three runs times two services by hand is thirty answers to read and record, once a month, forever. Most people do it once, find it depressing, and never do it again.
That is not a reason to buy anything. It is a reason to be honest with yourself about whether you will keep doing it.
Method two: the free check here
Four questions a couple in your city could type, asked to ChatGPT and Perplexity, result in about a minute, no card and no call. It tells you whether your name comes back and, when the service shows them, which sources the answer used.
It is deliberately smaller than the paid report. The report asks twenty questions built by reading your whole website, three runs each, on all four services, which is the two hundred and forty answers behind the numbers on this site. The free check is the version that answers one question: is there anything here worth looking at.
What to do with a bad result
Not panic, and not rebuild the website on Sunday night. A bad result on one service is normal, because the four do not read the same things. Across twenty measured photographers Gemini named a business on nineteen percent of questions while ChatGPT and Claude sat around ten, and Perplexity cited photographer websites two hundred and thirty times while Gemini cited them once.
So the first useful question is not what do I fix, it is which service am I missing on. If it is Perplexity, your own site is the lever, because Perplexity reads it. If it is Gemini, your own site is almost irrelevant and you need other people writing about you.
The longer version of what to do next is in how to show up when couples ask ChatGPT for a wedding photographer, and there are category pages for venues, planners and florists.
Questions people ask next
Why do I get different answers each time I ask ChatGPT?
Because the answer is generated fresh every time rather than looked up in a ranking. Small differences in what the service retrieves, and in the generation itself, are enough to change which four or five names come back. In one measured florist every single appearance across four services came in exactly one run out of three, and not one of them repeated. That is normal, and it is the reason a single test tells you almost nothing.
Does it matter if ChatGPT mentions me but does not link my site?
Being named is what wins you the enquiry, so it is the more valuable of the two. But being cited without being named is not nothing, because it means the service read your site and used it to build the answer, which is the step before it starts handing out your name. Perplexity in particular cites far more often than it names, so a check that only counts names throws away most of what it could tell you.
Should I use an incognito window when I test?
Yes, and log out. Assistants that know who you are can lean on your history, and if you have been reading your own website all week you are not testing what a stranger sees. It does not make the test perfect, but it removes the most obvious way to fool yourself.
How often is it worth checking?
Monthly, on the same questions, is the useful rhythm. Changes take roughly a month to show up because the services have to find and read the new pages first, and comparing this month to last month is the only way to tell whether something you did worked. Checking daily mostly measures noise.
Is the free check on this site really free?
Yes. Four questions, asked to ChatGPT and Perplexity, and you get the result in about a minute. No card. It is deliberately smaller than the paid report, which asks twenty questions built from your own site, three times each, on four services, but it is enough to tell you whether you have a problem worth looking at.