Discovery plan: understanding a launch decision

Decision to inform

Before adding templates or integrations, learn whether launch leads can explain why the fictional Analytics 2.4 launch is blocked and understand how new evidence changes a prior decision. The primary assumption is that an inspectable rule helps a readiness conversation; this has not been demonstrated with users.

Participants and approach

Recruit five consenting product operations or release leads who coordinate cross-functional reviews. Use only the seeded Northstar scenario. Ask about their review responsibilities without collecting confidential launch information. Ask permission separately before recording; participation and recording are distinct choices.

Start with a neutral task: “Describe whether this launch can proceed and the information you used.” Next ask the participant to prepare the missing sample information, record a decision, and explain what would happen if required evidence changed. Finally, show the sample incident and ask them to choose and explain a response. Do not introduce the interface as safer or more accountable before hearing their interpretation.

Assign two participants to describe their current checklist, tracker, or meeting process before the sample and three afterward. Record that order beside each session and use it only as a qualitative counterbalance, not an estimate of an order effect. Do not collect confidential launch artifacts. Participation and permission to record are separate choices, and notes should retain only the evidence needed for the stated study.

Evidence that could change the direction

Watch for participants treating the readiness percentage as permission, accepting arbitrary evidence text as proof, or mistaking a simulated action for deployment. Ask what remains uncertain, which information is unnecessary, and how their existing review document would serve the same task.

Record explanations, help requested, mistaken actions, and recovery. Repeated confusion should send the design back to the blocker explanation before feature expansion. If the existing process serves the task better, reconsider the need for a separate workspace. These five sessions are formative; the later quantitative targets are defined in the PRD and Validation plan. No research, recruitment, or participant outcomes are claimed here.