Exploration should not end with “the idea is validated.” It should end with a decision you can explain: what you learned, what remains uncertain, and whether the next investment of effort is worthwhile.

From findings to a decision

After exploring an opportunity, you may have interviews, observation notes, competitor lists, and market estimates. None of these automatically constitutes a conclusion. You must organize the evidence around questions: Is there a meaningful problem? For whom? How often does it occur, or what consequences does it have? What do people do today? How likely are you to reach them? What facts contradict your initial idea?

The SBA places customer, competitor, and market research at the beginning of business planning [R01]. GOV.UK recommends synthesizing discovery findings before defining solutions [R02]. This guide brings those principles together into an editorial decision framework: it is neither an official scoring system nor a guarantee of business success.

Step 1. Describe the opportunity without promising a solution

A useful description might follow this sequence: “In [context], [group] is trying to [task] but faces [obstacle]; the difficulty causes [consequence]. Today the task is handled through [alternatives]. We have observed [evidence], but still do not know [uncertainties].”

For example: “In restaurants that receive many reservation changes during peak hours, staff switch between several channels and sometimes lose information. We observed failures at certain establishments, while others manage the same process successfully. We do not yet know how common the problem is or which methods people would prefer.” There is no app in this statement because we have not chosen a value proposition.

Step 2. Summarize findings by dimension

Use a quick-reading table, not a report requiring someone to review fifty transcripts:

Dimension Decision question Evidence you might expect
Problem What specific difficulty occurs? Incidents, observations, records
Importance What consequences does it have? Time, loss, urgency, documented frequency
People Which segments experience it? Participants, contextual differences
Alternatives How is it handled today? Current methods, benefits, trade-offs
Market Are plausible scale and access present? Estimates with units, assumptions, and sources
Conditions What might constrain operations? Barriers, geography, rules requiring confirmation
Contradictions What challenges the idea? Negative cases, alternative explanations

If a cell is empty, do not invent its contents. An opportunity with several unexplored areas calls for a different decision from one supported by consistent observations.

When the central uncertainty is how many people you could serve, revisit market size and demand estimation.

Step 3. Identify assumptions that could change the entire decision

At this stage, you are not writing detailed experiments; you are spotting important dependencies. For a lunch service, these might be that a group has a recurring need, that you can reach it within the delivery area, and that current alternatives leave a meaningful problem unsolved. For a reservation system, they might be that the problem occurs in a recognizable segment and that switching does not introduce more work than the system aims to save.

An uncertainty becomes more urgent when its answer could force you to abandon or substantially redesign the idea and existing evidence is weak. Strategyzer proposes identifying critical hypotheses related to desirability, feasibility, and viability before designing tests [R10]. In this pillar, we make the uncertainty and its possible consequences visible; formal hypotheses and experiments will be developed in pillar 3.

Before treating a claim as well established, check its basis against the criteria for evaluating evidence and bias.

Assumptions capable of changing the decision should enter a critical-hypothesis priority map before resources are committed.

Step 4. Distinguish the problem from the conditions required for a business

A real problem does not establish a profitable solution. People's need for fast delivery tells you nothing about whether the delivery cost allows an acceptable price. A store's regular sales of a product do not prove that it would agree to change suppliers. You can describe a promising opportunity and simultaneously acknowledge commercial and operational doubts.

Distinguish three questions that later pillars will examine more deeply: desirability (is there a problem or need that matters to people?), feasibility (could we deliver a response using realistically available capabilities and resources?), and viability (could it sustain itself economically?). Strategyzer uses these dimensions to organize business assumptions [R10]. At this stage, you have neither a defined solution nor a full financial model, so a positive assessment remains provisional.

Step 5. Decide whether to proceed, investigate more, reframe, or stop

Proceeding to business design may make sense when the need is sufficiently defined, there are defensible signals within an identifiable group, and you know which unknowns must become proposals and tests. Proceeding does not mean labeling the idea “validated.”

Investigating further makes sense when a decisive question was not answered or the evidence conflicts. Specify the exact information that could change the decision; do not commission another generic study with no purpose.

Reframing is appropriate when the problem differs from what you expected, another segment experiences it, or the current alternative works better than anticipated. Sometimes the original solution ceases to make sense, but a more precise opportunity appears.

Stopping can be sensible if the problem is minor for reachable groups, existing alternatives work well enough, or known conditions make the idea impractical within its intended scope. Stopping early can save resources; document the finding that led to the decision.

If the decision is to proceed, opportunity findings become execution conditions; the execution planning guide shows how to turn decisions and unresolved assumptions into owned, verifiable deliverables, without pretending that research alone establishes operational readiness.

How to know when exploratory research is sufficient

No universal number of interviews, pages, or charts marks the end. Consider three practical tests: clarity, whether you can describe the problem, context, and people without hiding behind vague language; traceability, whether important findings have sources and limitations; and decision, whether you know what you will do with the information. GOV.UK's discovery guidance explains that the purpose is to understand different types of users and their needs before designing a service [R02].

If another similar interview is unlikely to change your problem definition but you have no idea whether people would switch alternatives, do not prolong the same interview indefinitely: it may be time to develop possible offers and design behavioral tests.

Decision example: appointments for small businesses

Research identifies three realities: some establishments with a high appointment volume lose changes during busy shifts; others manage paper diaries successfully; and several serve walk-in customers without needing a formal booking system. The problem is selective, not universal. A barrier also emerges: the manager worries about spending time migrating the diary.

A reasonable decision would be to redefine the target group around establishments with particular workloads and move to pillar 2 to compare value propositions, including process improvements as well as software. Another possible decision would be to observe more shifts if the evidence came from an exceptional day. None of the research performed yet establishes that the product will attract repeat buyers.

Deliverable: an opportunity profile and uncertainty register

The final profile should contain a summary of no more than one page plus supporting evidence. Include the problem definition, studied segments, alternatives, consequences and variations, preliminary market estimate, sources, contradictions, risks, decision and owner, and review date. Every important uncertainty should record “what we do not know,” “why it matters,” “what evidence we have,” and “what research or test might resolve it.”

You may also include options ruled out: “We did not prioritize occasional consumers because the organizational problem did not arise in the sessions available.” Preserve the limitation: those consumers may face different problems you did not study. This traceability helps the next team avoid reopening questions without knowing why decisions were made.

Common mistakes

Creating an opportunity score without explaining its dimensions; confusing verbal interest with intent to buy; choosing the largest population segment even though it does not experience the problem; treating an absence of competition as conclusive; passing financial assumptions off as observed data; and hiding negative evidence when presenting the project to others.

What you should have when you finish

A defensible opportunity description, a verifiable synthesis of evidence, a list of uncertainties prioritized by their effect on the decision, and an explicit next step: design the business, investigate a bounded question, reframe, or stop. That output feeds pillar 2; formal hypotheses and tests follow in pillar 3. If later evidence changes the picture, returning to pillar 1 is a normal part of learning.

For more detail, see types of business models and how to choose one and how to build a business model canvas step by step.

Main sources: [R01] SBA; [R02] GOV.UK, discovery; [R07] analysis; [R10] Strategyzer. The synthesis process and example are editorial proposals, not an official methodology.

If the decision is to investigate further, return to the step-by-step exploration guide with your updated research questions.

Bibliography and references