Consulting discovery hears 30 to 40 people, mostly senior, over six to eight weeks. AI interviews hear everyone, at every level and in their own language, in about three weeks, with the same probing in every conversation. What does not change is the judgment: the recommendation, the politics and the change programme stay human work.
The old constraint
A senior consultant can hold four or five good interviews in a day. That single fact shaped everything about the traditional diagnostic: the sample of 30 to 40 people, the discovery phase of six weeks or more, and a quiet bias toward the people who are comfortable talking to consultants, which usually means the same senior voices leadership already hears.
None of this was anyone’s preference. It was arithmetic. Every experienced partner knows that forty interviews in a company of five thousand is a thin slice. The model simply never allowed more.
The comparison, directly
| Compared | Consulting discovery | AI interviews |
|---|---|---|
| Coverage | 30 to 40 people, senior-skewed | Everyone, every level |
| Time to findings | 6 to 8 weeks | About 3 weeks |
| Languages | Whatever the team speaks | 32, by voice or chat |
| Honesty | People perform for the room | No one to impress, anonymous by design |
| Consistency | Varies by interviewer and by day | Every conversation probes the same way |
| Measuring again | A new engagement | Same people, same scoring |
What AI interviews do not replace
Judgment. The recommendation, the political navigation, the board conversation and the design of the change programme remain human work, and they are the most valuable part of what good advisors do. What changes is the foundation under that judgment. Instead of reasoning from a sample, the thinking starts from the whole company.
This is why advisors are among the firms that work with us: the machine does the listening, their people do the thinking, and the discovery phase stops eating the engagement.
“The machine does the listening. People do the thinking.”
The honesty effect
The least expected result of interviewing at scale is candour. With no career risk in the room, no face to read and no one to impress, people describe where the work breaks in specific, fixable detail. Frontline employees who would never contradict a consultant in a meeting room will tell a private interviewer exactly which step gets skipped on a busy shift, and why.
The bottom line
“AI interviews or a traditional diagnostic” is the wrong question. The better one is what you could know about your company if listening were nearly free, and what it costs you to keep not knowing it. For twenty years, sampling was the ceiling. The ceiling is gone.
Questions
Do we still need consultants?
For the judgment, yes. AI interviews replace the listening and the synthesis, not the decision or the change programme that follows.
Is a whole-company study slower to read than 40 interviews?
No. Every conversation is broken into checked claims and rolled up into ranked findings, so leadership reads a short set of battles, not thousands of pages.
Can we compare it with a diagnostic we already ran?
Yes. Pick a question you already paid to answer, ask the whole company, and compare what comes back.
Nour, our guide, will suggest the right study and show you what it would find. Five minutes, any hour.
Talk to Nour