Skip to main content

Step 10: Build your hypothesis document

5 XP

Phase 1: Preliminary Research · Step 10 of 13

Research becomes strategy when it is organized into explicit, testable beliefs. Your hypothesis document is the foundation that every subsequent decision builds on.

Article​

A hypothesis document is not a business plan. A business plan is a narrative designed to persuade. A hypothesis document is a structured set of beliefs designed to be tested. The difference is critical. Business plans resist disconfirmation. Hypothesis documents invite it.

Each hypothesis in your document should have a clear structure. The belief states what you think is true. The confirmation evidence describes what you would see if the hypothesis is correct. The disconfirmation evidence describes what you would see if the hypothesis is wrong. The confidence level reflects how much evidence currently supports it. The criticality rating tells you whether the business survives if this hypothesis fails.

The discipline of writing falsifiable hypotheses changes how you think about your business. Instead of collecting evidence that supports your idea, you start looking for evidence that would disprove it. This is uncomfortable but enormously valuable. Every hypothesis that survives a genuine attempt at disconfirmation becomes a stronger foundation for your strategy.

You should produce between five and ten hypotheses. At minimum you need hypotheses about your customer identity, their problem, why existing solutions fail them, their willingness to pay, and the market size. If your research from the previous steps is insufficient to form a hypothesis on any of these topics, that is a finding. It means you need more research before you can proceed.

This document is alive. It will evolve as you conduct customer interviews, gather survey data, and test your assumptions in the real world. The version you produce now is your starting point, not your final answer. But it needs to be clear enough and honest enough to be a useful starting point.

Key Takeaways​

  • Why: If you cannot state your hypothesis in a way that could be proven wrong you do not have a hypothesis you have an assumption.
  • What: Use AI to synthesize your research into a structured list of falsifiable hypotheses.
  • Output: A hypothesis document with five to ten explicit beliefs about your customer, market, and opportunity, each with a falsifiability condition.

AI Prompt​

Act as a strategic analyst. Based on all of my research from the previous steps: [paste your validated research summary from step 9 here].

Build a hypothesis document for my business opportunity. For each hypothesis:
- State the belief clearly in one sentence
- Identify what evidence would prove this hypothesis correct
- Identify what evidence would prove this hypothesis wrong
- Rate your current confidence as high, medium, or low based on the research quality
- Identify whether this hypothesis is critical meaning the business fails if it is wrong or supporting meaning it affects strategy but not viability

Produce between five and ten hypotheses. Cover at minimum: who the customer is, what problem they have, why existing solutions fail them, whether they will pay for a better solution, and whether the market is large enough. Do not add hypotheses that are not grounded in the research. If the research is insufficient to form a hypothesis on a topic say so explicitly.