Entrepreneurial Opportunity, Problem and Customer · Lesson 4

Hypotheses, Assumptions and Opportunity Screening

5 min read

Lesson objectives

Turn an assumption into a hypothesis with behavior, a test, and a predefined decision threshold.
Prioritize assumptions by impact if wrong and uncertainty to decide what to learn first.
Evaluate evidence strength and inference limits and choose whether to continue, modify, or pause.

Turn an assumption into a hypothesis that can lose

An assumption is a belief used in a decision before it is sufficiently supported, such as “Tutors see rescheduling as an important problem.” A hypothesis makes that belief more testable by specifying for whom, in what context, what behavior is expected, what test will be used, and what result would strengthen or weaken the belief. A useful hypothesis is not written so that every result counts as success; it must allow a clear possibility of learning that we were wrong.

A practical format is: “For [segment] in [context], we believe [observable behavior or outcome]. We will examine it with [test], and we will use [a predefined decision threshold] to decide whether to continue, modify, or stop.” The threshold is not a universal scientific truth; it is a stage-specific decision rule that should be justified before seeing the result so the goalposts are not moved afterward.

Test the riskiest assumption, not merely the easiest

Assumptions are not equally important. A practical prioritization uses two lenses: impact if wrong and current uncertainty. If an assumption being false would make the venture unreasonable and the evidence for it is weak, it is a strong candidate for early testing. A low-impact, low-uncertainty assumption may wait. This is a thinking matrix, not a scientific equation; the point is to explain why a particular error could change the decision rather than testing convenient details that do not matter.

Match evidence to the claim

The question “Does the problem exist?” may benefit from interviews about past events. “Will the user complete a particular workflow?” may require a prototype or usage test. “Will the customer actually commit?” may require a stronger signal such as booking time, entering real non-sensitive data, placing a trial request, or taking a genuine purchase step when ethically appropriate. No single evidence type is best for every claim; evidence is stronger when it is close to the behavior or outcome being judged.

Approaches such as Lean Startup can make iterative learning and experiments more explicit, but they are tools, not guarantees of success or complete scientific truths. Peer-reviewed work discusses how experimentation can reduce uncertainty while also documenting limits around hypothesis generation, learning from customer feedback, and the possibility that experiments favor incremental improvements over broader novelty. We therefore use a “hypothesis — test — learn — decide” cycle with contextual judgment rather than as a ritual to follow mechanically.

Exercise

Exercise 1 — Turn three assumptions into hypotheses: “the problem matters,” “people will use the solution,” and “the price is acceptable.” For each hypothesis, specify the segment, expected behavior, test, a predefined decision threshold, and the result that would trigger reconsideration. Make the threshold appropriate to the learning stage and do not present it as a universal rule.

Check your answer

Concise guided solution for problem importance: “Among independent tutors with frequent changes, we expect most participants who meet the criteria to describe a recent situation where rescheduling caused tangible rework. We will use interviews about the previous two weeks and define rework in advance. If examples are rare or unrelated to schedule changes, we will reframe the problem.” Usage requires a behavioral test, while price requires evidence closer to a purchase decision than merely asking, “Is this price acceptable?”

Exercise

Exercise 2 — Write a six-line opportunity screen for FlexSlot: problem, segment, potential value, strongest current evidence, riskiest unsupported assumption, and smallest next test. Then write a three-way decision: continue, modify, or pause, with a condition for each outcome.

Check your answer

Guided solution: “Problem: repeated appointment updates across multiple tools when schedules change. Segment: independent tutors who personally manage frequent changes. Potential value: less rework and confusion. Evidence: a limited set of cases described the behavior; there is no market estimate. Riskiest assumption: the problem matters enough to change the current method. Next test: recruit more diverse interviews, then test a simple workflow if the pattern persists. Continue if evidence strengthens, modify the segment or problem if a different pattern appears, and pause if the problem remains weak relative to the current alternative.”

Lesson summary

A useful hypothesis can be wrong and connects a claim with a test and a predefined decision threshold. Test high-impact, high-uncertainty assumptions first, match evidence to the claim, and do not turn small samples or entrepreneurship tools into universal rules of success.