Case: Harvey launched an AI contract-review agent for enterprise legal teams, but adoption is lower than expected. How would you diagnose the cause, define success metrics, and decide what to do next?
My answer mainly revolved around:
- Adoption funnel: eligible → exposed → activated → completed workflow → retained
- Output quality: whether the results are accurate, usable, and how much manual editing they need
- Customer value: whether it saves lawyers' time and makes contract review more efficient
- North Star: workflows that are completed successfully and produce a usable legal outcome, rather than just raw feature usage
Discussion
Loading comments…