Give the number, then give the range, then say what decision each end of the range implies. Executives are not allergic to uncertainty, they are allergic to unusable answers. Framing it as forecast 12 percent growth, plausible range 8 to 17, and here is what we would do differently at each end, turns uncertainty into something actionable rather than a hedge.
Why interviewers ask this
Communication questions for data scientists are really about whether your rigor survives contact with the business. The interviewer wants to see that you neither hide the uncertainty to look confident nor bury the answer under caveats. Tying each end of the interval to a different decision is the technique that shows you understand what the executive is actually trying to do with your number.
How to structure your answer
- Lead with the point estimate so the question gets answered.
- Follow immediately with the range in plain units.
- Attach a decision to each end of the range.
- Say what would narrow the range and what that would cost.
Example answer
I lead with the number, because refusing to give one reads as evasion and they will just get it from someone less careful. Then I attach the range in the same breath. Something like: our best estimate is 12 percent growth next quarter, and the plausible range is 8 to 17. The part that makes it land is what comes next, which is tying the ends to decisions. At 8 percent we hold headcount; at 17 we need two more support staff hired by March. Now the uncertainty is not a caveat, it is a contingency plan, and that is a conversation executives are comfortable having. I also say what drives the width. If the range is wide because we only have six months of history on a new segment, I say so, and I say what would narrow it and roughly what getting there would cost. What I try never to do is present a single number as if it were certain, because the first time reality lands outside your implied precision, everything you say afterward gets discounted.
Walking into this interview soon? GhostPilot listens to your live call, spots the question the moment it is asked, and puts a structured answer on your screen in real time. Try it on your next mock, or grab a $29 Session Pass, no subscription, for the real thing.
See how it worksFollow-up questions to expect
- What if they push back and insist on one number with no range?
- How do you present uncertainty in a slide rather than out loud?
- How would you handle a forecast that turned out badly wrong last quarter?
Related data scientist questions
Your interviewer will ask their own version of this. Paste your actual job description into the free Question Predictor and get the 20 questions that role is most likely to ask, with what each one is really probing.
Predict my questions