ORIGINAL COMPANY-CONTEXT PRACTICE
Measure product adoption when accounts and sites differ
Original practice scenario: A hypothetical medical-technology analytics team measures new-device adoption across hospital systems. Each customer account owns several sites and each site may have multiple devices. Define useful cohort metrics.
Written for independent practice using company context; this is not a verified question from a company interview.
A REASONED APPROACH
Model answer.
Choose whether adoption means account, site or device activity. These grains answer different questions, and a raw join can make one hospital system appear to adopt repeatedly.
Work through it
- Define an eligible population and a first qualifying activation event, excluding test units and retaining the applicable account-site relationship at event time.
- Build one row per chosen cohort unit, then calculate activation and later activity with explicit observation windows and a consistent time zone.
- Report counts at account, site and device grain separately, preserving unknown ownership mappings and sites too recent to complete the follow-up window.
Tradeoffs to discuss
- Account-level metrics describe customer reach while device-level metrics reflect deployment depth; a single percentage cannot replace both.
Common pitfalls
- Do not count administrative transfers as new adoption without an explicit rule.
- Distinguish missing activity telemetry from a confirmed inactive device before labeling retention.