A B2B software company's support tickets rose from 1,200 to 2,500 a month. No release shipped in the period. Active accounts grew 6%. The increase is almost entirely 'how do I' questions rather than bug reports. Tickets from accounts older than a year are flat; the rise is in accounts onboarded in the last four months. The company changed onboarding three months earlier from a live 90-minute session to a self-serve video series, to cut cost. Activation rate — accounts completing setup within 14 days — is unchanged at 71%.
Isolate the problem, give the causes worth testing, say which the evidence supports, and recommend a next step.
100 points, 60% to pass.
The segmentation is decisive: old accounts flat, new accounts driving it, and the onboarding change lands in the same window. Activation being unchanged is the trap — it says people still finish setup, so the change looks successful on the metric it was judged by, while the cost has moved downstream into support. Strong answers name that the saving was not a saving but a transfer, and quantify it: 1,300 extra tickets a month against the cost of the live sessions removed. The next step is not restoring the old onboarding wholesale but finding which parts of the 90 minutes the videos failed to carry, which the ticket topics will show.