Innovation Accelerator Hub
All playbooks
Operating model stages

Discovery

Frame the problem, map the users, and shape a hypothesis sharp enough to prototype against.

Published
Owner · Director, InnovationUpdated · 2026-07-19

Purpose

Discovery turns a pursued signal into a hypothesis sharp enough to prototype against. The goal isn't a research report — it's a testable statement of who, what, and what "better" looks like.

Entry
Pursued signal with a hypothesis owner.
Artifact
Problem frame + user map.
Gate
Testable hypothesis.

Activities

  • Frame the problem. Job to be done, current workaround, cost of the workaround, moment of pain.
  • Map the users. Who has the problem, who benefits, who blocks change, who owns the metric.
  • Two to four user conversations. Not interviews for a report — conversations that pressure-test the frame. Record them.
  • Write the hypothesis. "We believe [user] will [behavior] because [reason], and we'll know because [signal]." One sentence.
  • Pre-write the prototype prompt. If you can't describe what to build in a paragraph, the hypothesis isn't sharp enough yet.

Common failure mode

Discovery becomes a deliverable
Discovery is not a phase for producing decks or Figma boards. If you're spending more than a week here without a testable hypothesis, you're building the wrong artifact. Prototype the hypothesis you have and let the prototype pressure-test the framing.

Supporting tools

  • Recorded sponsor and user conversations.
  • ChatGPT to synthesize transcripts into a first frame.
  • Prototyping workflow for the immediate next step.