SectorsAI findability for professional services
AI findability for professional services
Your clients ask ChatGPT before they call you. Whoever gets named there is in the conversation before it even starts.
What it usually gets called:Findable in ChatGPT, AI SEO, GEO, answer engine optimization
The question
Why does ChatGPT name other advisory firms and not yours?
Because an AI assistant only names what it already knows. An answer is not an opinion but a summary of a handful of pages the model trusts at that moment, and those pages rarely sit on the website of the company being named. Whoever does not appear in those sources does not appear in the answer. So this is not about writing better copy on your own site; it is about appearing in the material the answer is assembled from.
For a professional services firm that coincides with the moment the choice gets made. The question of which firm could help with this now goes to an assistant first and to your own network second. The names that come up decide who gets called at all. A knowledge centre on your own site helps less with that than it seems: it is where you record what you know, but it is not where the model gets its answer.
The pattern · 28 August 2026
Where those names come from
We counted where those sources come from. By far the largest group was pages by other agencies writing about the same craft. For a firm with its own knowledge centre that is the uncomfortable part of the outcome: the answer about your market is written by parties in your market, and that happens whether you take part or not.
244 mentions · 47 sites
76 mentions · 15 sites
50 mentions · 4 sites
This breakdown comes from our own measuring round on the market for AI findability, not from a measurement in this sector. No per-industry breakdown exists; what it shows is how an AI answer gets assembled, not what comes out in your market.
How we measure it
One number, and what it consists of
We summarise the outcome in one number, the TAZM score: which share of its own search market a company owns, on a scale of 0 to 100. It is the average of four components, of which only the ones actually measured count, and every number carries its numerator, its denominator and its measuring date.
The limit
What we do not promise here
We do not promise you a place in an AI answer. It cannot be bought and nobody can guarantee it: that is precisely the question the measurement is about. What is settled is the measurement itself: the same questions, the same platforms, the same measuring date, every round again, and we publish a disappointing outcome just as readily as a good one.
Concrete
A client asking which firm could help them with this, and getting three names back that do not include yours
The second layer: the systems behind it
Findability brings the question in. What happens after that is a second problem, and for professional services it nearly always has the same shape.
In a firm with a handful of partners, judging a new case sits in the head of whoever has been there longest. That works, right up until that person is away for a week. The cost is not in the mistakes but in the waiting: every file sits still until the right person looks at it.
What a system takes over here is the first assessment, not the judgement itself. The partner approves, the system does the groundwork. The criteria come out of the decisions already made: what the partner recognises as his own reasoning gets recorded, what he does not recognise does not go in.
That side is written out at greater length on our professional services page, with the questions that keep coming back there.
Common questions
- We rank well in Google. What would AI findability add?
- They are two different audiences. A Google position brings someone to your site; an AI answer summarises the market before anyone clicks. In 28 of the 30 searches we measured in our own market on 28 August 2026, an AI answer sat above the results. Whoever is not named in it loses the question before the position underneath matters.
- We have published a knowledge centre for years. Is that not enough?
- It is the precondition, not the outcome. Without pages of your own there is nothing to cite. But in our own measuring round of 28 August 2026, most of the sources that came back systematically were not company websites. Publishing and being cited are two steps, and the second one is the work.
- Can you guarantee that we will appear in ChatGPT?
- No, and neither can anyone else. What we do record is the measurement: which questions, which platforms, which measuring date, and what actually came back. That list of questions is settled before the first measurement and is not narrowed afterwards, so a repeat measurement can be compared with the first.
AI findability for professional services: where do you stand today?
Measure first, choose after. The AI Findability Scan records your findability map and your TAZM score, with numerator, denominator and measuring date. The scan page itself is in Dutch.