The Research Sprint

A research sprint is a structured thirty-minute block that takes you from question to deliverable. It is not a casual exploration — it is a disciplined process with four phases, each with a time budget. The sprint format works because it imposes urgency. Without a time constraint, research expands to fill whatever time you give it. With a thirty-minute box, you are forced to prioritize, make decisions, and produce output.

1. Phase 1: Question (5 minutes) Open your AI conversation with the research brief. Use the role instruction technique from Lesson 5. Set the scope, define the deliverable format, and ask your first question — the broad landscape question from the question ladder.
2. Phase 2: Explore (10 minutes) Work through the question ladder: broad to specific to actionable. Use the follow-up techniques from Lesson 6 — challenge assumptions, request evidence, explore contradictions. Cover the terrain, but do not get stuck on any one thread.
3. Phase 3: Verify (5 minutes) Apply the verification habit to your key findings. Check the two or three claims that your conclusions depend on. Ask the model which claims it is least confident about. Cross-reference anything load-bearing.
4. Phase 4: Synthesize (10 minutes) Ask the model to synthesize the conversation into your specified deliverable format. Then refine: "Make the recommendations more specific," "Add confidence levels to each finding," "What did we not cover that a domain expert would want to see?"

The sprint format is deliberately rigid for beginners. As you build the research habit, you will naturally adjust the time allocations — spending more time on exploration for novel topics and more time on synthesis for complex ones. The structure is scaffolding. Once the habits are automatic, the scaffolding comes down and you research fluidly. But start with the structure. It prevents the two most common beginner mistakes: exploring without synthesizing, and synthesizing without verifying.