Estimation question

Estimate the delta in the sale of hair trimmers/clippers during the COVID-19 lockdown.

Practice this question out loud. An AI interviewer asks it, follows up like a real interviewer would, and scores your answer. Type or speak.

What this question tests

Estimation reasoning about a behavioral shift caused by a specific real world event, requiring a before-and-after comparison with clear logic.

State the hypothesis directly: lockdowns closed barbershops and salons, likely increasing demand for at-home hair trimmers and clippers.
Break the estimate into a funnel: baseline pre-COVID monthly trimmer sales, times an assumed multiplier for the surge during lockdown months.
State assumptions explicitly, for example baseline monthly sales in a market, and a surge multiplier, perhaps 2x to 4x, reasoning that barbershop closures pushed a large share of regular haircuts to at-home solutions.
Consider offsetting factors: some people simply let hair grow rather than buying a trimmer, which would reduce the multiplier from an extreme high estimate.
Calculate a delta (increase in units or revenue) and sanity check it against the fact that this was a widely reported, real consumer trend during 2020 lockdowns.
Segment by whether this was a one-time spike (early lockdown panic buying) versus a sustained shift, since the delta likely tapered as salons reopened.

States the surge multiplier as an explicit, reasoned assumption rather than an arbitrary number pulled from nowhere.
Accounts for an offsetting behavior, people simply growing their hair out instead of buying a trimmer, showing balanced reasoning.
Distinguishes a temporary spike from a sustained shift, correctly reasoning that the delta likely shrank once salons reopened.

Giving one flat multiplier for the entire lockdown period without considering it likely tapered over time.
Ignoring substitute behaviors, like growing hair out, that would reduce the actual delta.

How would this estimate differ across regions with different lockdown lengths?
How would you validate this estimate using actual retail sales data if you had access?