When Estimation Beats Analysis

Under time pressure, a fast rough estimate often beats a precise calculation that arrives too late. The point is not that estimates are more accurate — they are not — but that a good-enough answer now can be worth more than a perfect answer next week.

Many consulting firms drill new hires in Fermi estimation for exactly this reason. A classic case-interview prompt: in two minutes, estimate the daily revenue of Starbucks in Manhattan. One path — about 8 million people in NYC, roughly 20% in Manhattan (1.6M), maybe 1 in 10 buying daily at about $5.50 each — lands near $880,000/day. Rough, but the right order of magnitude.