The questions we ask in the first 30 minutes
A working list of the questions we use to figure out whether an AI engagement is worth either side of the table starting.
Most intro calls are a sales motion. Ours aren’t. We’ve turned down a lot of work that wasn’t a fit — for them, for us, or for both — and we’ve learned to figure that out fast.
These are the questions we actually ask, in roughly the order we ask them.
1. What are you trying to be true a year from now?
Not “what AI tools are you evaluating.” If the operator can answer this clearly, the AI question gets a lot easier. If they can’t, no tool will save them.
2. Who on your team would own this if it worked?
If the answer is “we’d hire someone,” that’s a different engagement than “Sara would run it.” We can build either, but the cost and the timeline diverge fast.
3. What did you try last year that didn’t work — and why?
Two reasons we ask: one, it tells us what’s been ruled out. Two, the answer reveals how the company actually decides things, which matters more than the answer itself.
4. What would have to be true for you to kill this in 90 days?
If the operator can’t articulate a kill criterion, they’re not ready. The willingness to walk away is what makes the bet worth taking.
5. Who else is in the room?
Single decision-maker engagements move 4x faster than committee engagements. We can do both. We charge accordingly.
We ask these because the answers tell us whether we should be working together — not because we’re going to write them up in a report. The report is the work.