How we work
We never pass off noise as a result.
On Monday, we ask ChatGPT the same question 10 times: the brand is mentioned 4 times. On Tuesday: 6 times. Nothing has changed. In a report, that is written up as +50%.
How we work
No promise without a measurement.
Measure
A baseline in numbers, before anything else.
Prepare
We get your data in order with you. Ready to grow.
Propose
Build, buy or make better use of what you have.
Prove
Measured again afterwards. Not a gut feeling.
Our protocol
-
1. We repeat, we vary
Each question, 10 times, to each AI assistant: it never answers the same way twice. The more different questions we ask, the more precise the figure.
-
2. We check
Every measurement is repeated on another day. If the two disagree, we draw no conclusion. And a person reviews answers drawn at random, to check that the machine reads them correctly.
-
3. We prove
Every figure comes with its range (a 95% confidence interval), which shows how precise it is on the questions asked, and links back to the answers we kept.
Result: between two dates, a real gap of 10 percentage points is detected more than 9 times out of 10 (in simulation, with the full protocol).
Example measurement
Mentioned by ChatGPT
Mentioned by Gemini
Demonstration measurement on a recent film, not client work, and shorter than a campaign: rough figures, with no range. Film and competitor names anonymised.
[n] questions, asked [n] times, to [n] AI assistants
Our one limit
The “before” measurement has to be taken before: it cannot be rebuilt afterwards.
Let’s talk about what’s taking up your time.
A 30-minute video call, booked in our calendar.
Contact us