Top 7 Forecasting Interview Questions
Browse 7 real-world technical and behavioral interview questions about Forecasting. Review scenarios, edge cases, and architectural best practices.
An executive wants a firm date for work your team has not scoped yet. What do you do?
Reject the choice between a fabricated date and no answer. Ask what decision the date serves, separate known work from assumed and unknown work, give a range with its confidence and assumptions attached, buy the missing information with a timeboxed spike, and offer to fix scope or date but not both.
There are forty items left before the release and over the last twelve weeks the team has finished between four and nine a week. The sponsor wants a date. What do you tell them?
Forecast from the whole spread of observed weekly throughput rather than its average, subtract the rate at which the backlog is growing, and give a date with a confidence attached. Then say plainly whether you are offering a forecast or accepting a commitment, because they are not the same promise.
You forecast every store and also the national total, and the store forecasts do not sum to it. What do you do?
Forecast at whichever levels have signal, then reconcile so the numbers are coherent by construction. Bottom-up preserves detail but accumulates noise, top-down is stable but allocates by an assumed proportion, and optimal reconciliation projects independent forecasts onto the set that adds up.
Leadership has asked for 20 per cent growth in activations next quarter. How do you decide whether that is the right target?
Build the number from a baseline and named mechanisms rather than accepting it, because a target with no route attached is a wish that will be met by lowering the definition. Decompose the growth into drivers, state what each is worth, and say plainly which portion of the gap has no plan behind it.
Which delivery metrics would you actually track for a team, and why does velocity stop working the moment you set a target for it?
Track throughput, cycle-time distribution and flow efficiency, because they describe the system rather than the estimate. Velocity is denominated in a unit the team defines, so making it a target lets the number rise without any change in delivered output, and it stops measuring anything.
Why does a random train-test split ruin a forecasting model?
A random train-test split ruins a forecasting model by leaking future observations into training. Use time-based holdouts or rolling-origin backtesting, and build every feature only from data available before the forecast timestamp. Use this time series answer to show the decision, trade-off, and evidence rather than a memorised definition.
Why do we estimate in story points at all, and what would you do if the team's estimates were consistently wrong?
Points are relative sizes that trade false precision for a fast conversation whose real output is surfaced disagreement. Estimates that are always wrong are a system symptom, so diagnose unclear criteria, oversized stories, invisible work and interruptions, then forecast with throughput and cycle time.