Run a Pre-Mortem Before Shipping the Agent
Use a pre-mortem to identify likely agent rollout failures and convert them into design changes before launch.
The team plans to launch a support agent next week after a successful pilot, but the operating risks have not been stress-tested openly. Pre-mortem = imagine failure -> list causes -> rank likely/high-impact causes -> convert each into a control The common trap is waiting for confidence before asking failure questions. That delays the useful skepticism until after the blast radius has grown. Assume failure State the premise: “It is 30 days after launch and the rollout went badly.” This removes the social pressure to sound optimistic and invites honest risk surfacing. List causes Ask each stakeholder to write why…
Sign up free — one personalized lesson every day, matched to your role and goals.
Already have an account? Sign in