Anyone who wants to know whether their business is mentioned in ChatGPT, Perplexity or Google's AI Overviews needs a measurement instrument. The choice is between four specialised tools and a spreadsheet with discipline. Which of the five fits depends on how many questions you track, how many brands you measure and which platforms matter for your clients.

01 / the instrumentWhy you choose a measurement instrument first

Without a fixed measurement, any statement about AI visibility is an opinion. Answers differ per session and per phrasing, so only a fixed question list that you repeat at fixed intervals produces an outcome that still tells you something next month. How that benchmark works is set out in how you measure AI visibility; this piece is about the tools you use to run that measurement.

To be clear: this is a young corner of the field. In the Netherlands, this category is still searched for relatively little (according to NL search data from DataForSEO, August 2026, about 40 to 70 searches per month for terms like ai visibility tool and ai seo tools), and the tools themselves are young. So expect coverage and capabilities to shift, and check with the maker what applies today.

02 / the comparisonFour tools and the spreadsheet side by side

The differences aren't in what the tools promise, but in which platforms they read and what their reporting focuses on. This is the coverage as the makers publish it (as of August 2026):

We deliberately leave out prices: they change quickly in this segment and weren't in our source. And again the honesty that a comparison calls for: we haven't tested any of the four in production over a long period. What's here is the published coverage, not a practical verdict.

03 / the zero optionWhy manual work is a valid choice

The spreadsheet isn't a poverty version of a tool, but the same measurement without a subscription. You fix your question list, open each platform, note per question whether you appear, and repeat that every month at the same time. How you set up such a first round in half an hour is set out in how you check whether your business appears in ChatGPT.

A tool speeds up your measurement; the benchmark is what makes it reliable.

What manual work doesn't give you is scale and history without effort. Twenty-five questions across four platforms is half a day per month; two hundred questions across five platforms for three brands is no longer manual work. That's where the tools' territory begins.

04 / the choiceWhere you start, and when you switch

Our advice is simple: start with manual work and a fixed benchmark, and only switch to a tool once the number of questions, platforms or brands makes the manual work unsustainable. That way you first learn what a meaningful question and a meaningful outcome are, and afterwards you buy with purpose rather than on the promise of a dashboard.

If you do choose a tool straight away, choose based on the platforms where your clients search, not the longest feature list. A B2B audience that works a lot with Claude and Copilot gets little from ChatGPT-only coverage; anyone who mainly needs to be found on Google looks first at AI Overviews coverage.

When choosing, pay attention to two practical points, whatever brand is on the box. First: can you export your measurement series, so your history moves with you if you later switch or return to the spreadsheet. Second: can you enter your own question list instead of a generated one, because the measurement is only comparable to your baseline measurement if the questions stay the same.

And for anyone who'd rather not run the measurement themselves at all: that's also available as a service, including a report and improvement points. What such an outsourced baseline measurement involves is set out on the scan page. For anyone whose clients don't come in through online search or AI answers, the reverse applies: none of these five options is worth your time, and that's the most useful outcome of this piece.

This text was produced with AI assistance and checked and approved by a human before publication.