Who does AI mention in our own market?
In our own market, DataDrift Digital sits at around fourth place, and yet even the market leader appears in only 11 of 100 AI answers. Our own measurement, 29 September 2026, 780 answers from four AI assistants.
Lees dit in het Nederlands →In our own market, DataDrift Digital is among the providers that AI mentions most often: around fourth place (range 2 to 12), a little more than half as often as the market leader, and more often than each of the three other companies we scanned, each in their own market. Still, that says something notable about AI: even the market leader appears in only 11 of 100 answers. Even those who score well are mostly not mentioned. That's where the room is.
We measured this ourselves on 29 September 2026: 65 buying questions, each asked 3 times to ChatGPT, Gemini, Claude and Perplexity. We appear in 6 of 100 answers, the market leader in 11.
We measure for others how often AI mentions them. Our own result belongs in that too, including the weak spot.
How we measured
- Measurement date: 29 September 2026. The next measurement is on 4 January 2027.
- Questions: 65 buying questions. Fifty are general, fifteen describe the asker's own company. Questions containing our own name do not count.
- Assistants: ChatGPT (gpt-5.5-2026-04-23), Gemini (gemini-3.6-flash), Claude (claude-sonnet-5) and Perplexity (sonar), via the API, with web search on.
- n: each question 3 times to each assistant: 780 answers. An answer counts if it mentions DataDrift Digital by name.
- Comparison: the market leader is the competitor AI mentioned most often in this measurement, on its own denominator: questions containing its name do not count, so 756 answers.
- Prepared by: Pim van den Boomen. Drafted with AI, checked by Pim van den Boomen.
Four examples of the questions:
- Which agency in the Netherlands can help me become visible in ChatGPT?
- Where do AI assistants get their information about companies from?
- What is the difference between being found on Google and being mentioned by AI?
- Which party helps with AI visibility and then measures whether it worked?
The result
| What we measured | DataDrift Digital | The market leader in our market |
|---|---|---|
| Mentioned per 100 answers | 5.9 (46 of 780) | 11.4 (86 of 756) |
| Assistants that never mentioned us | 1 of 4 (Claude) | n/a |
| Buying questions on which no assistant mentioned us | 71 per cent | n/a |
| Position among the providers AI mentions | around fourth place, with a range of 2 to 12 | 1 |
So the market leader is mentioned 1.9 times as often as we are. Our proof page also shows a TAZM score of 13.3: that is a composite score from an earlier measurement with different questions (9 September 2026), not a share of answers, whereas 5.9 per 100 counts how often an AI answer mentions us by name.
What does that mean for you? You don't have to be number 1 to count: even the market leader is absent from almost 9 of 10 answers, and with 6 of 100 we still sit at around fourth place. But whoever AI never mentions does not exist in those answers. On 71 of 100 buying questions, no assistant mentioned us, not even after asking 3 times.
Is this different for others?
We carried out the same measurement for three other Dutch SME service providers, each in a different market, measured on 29 September and 6 October 2026. Of the four companies we scanned:
| What we measured | DataDrift Digital | Median of the four companies we scanned |
|---|---|---|
| Mentioned per 100 answers | 5.9 | 2.5 |
| Buying questions on which no assistant mentions the company | 71 per cent | 89 per cent |
| Answers that do mention a provider from the same market | 55 per cent | 50 per cent |
In 4 of the 4 companies we scanned, the scanned company was not the market leader, and Claude did not mention the company in any scan. In roughly half of the answers, AI does mention someone from the same market. The question gets answered, just with a different company.
What this measurement does not say
- One measurement moment. A measurement date, not a trend. Only the remeasurement on 4 January 2027 will show whether anything shifts.
- Via the API, not the app. What someone sees in the app can differ: personal history, location, logged in or not.
- Location. 53 of the 65 questions do not mention a location, 12 mention a location in the south of the Netherlands.
- Claude searched less. Claude also searched the web, but cited sources less often and mentioned a company less often. So a zero for Claude is partly a feature of Claude.
- Four companies is not a market average. It is a pattern in our scans.
- No names. We do not name the market leader or the other companies.
Where do you stand?