Quantify database bug cost/benefit

A registration defect impacts Premium members during the period spanning January 1 through March 31, a three-month window. To qualify for the 2% commission, a Premium user needs to be registered; those not registered receive the 1% commission rate; the monthly administrative fee of $1.25 is charged solely when registration is active. The bug produces two effects: first, Premium users who cancel their accounts stay registered; second, newly joining Premium users are left unregistered. On a monthly basis, 400 Premium users cancel, while 2,000 new Premium users sign up. Take average monthly spending as $625 for Premium and $500 for Regular. Commission counts as revenue for the company; the admin fee is a cost to the company. These cohorts continue until the fix is applied on April 1 (for instance, January's cancelers stay registered from January through March; February's from February through March; and so on). Calculate the net profit effect across the three months compared to proper operation, breaking down the revenue and cost differences for each cohort and providing the overall dollar amount.

Overview: This question assesses a candidate's skill in conducting cohort-based financial impact analysis and applying basic statistical arithmetic to measure the revenue and cost changes resulting from registration-state mistakes.

Read the complete Capital One Data Scientist interview experience that featured this question.