Opportunity Framing and Assumption Mapping

Introduction

Entrepreneurial opportunities rarely arrive as fully formed business ideas. They usually begin as observations: a recurring inconvenience, an underserved group, a new technology, a regulatory change, or an inefficient transaction. The central task of customer discovery is to transform such observations into propositions that can be investigated. This requires more than creativity. It requires disciplined framing of what is believed, why it is believed, and what evidence could prove the belief wrong.

Opportunity framing is the process of expressing a proposed venture in terms of a specific customer, problem, solution, channel, revenue mechanism, and growth mechanism. Assumption mapping then separates what is known from what is merely believed and prioritizes the beliefs that pose the greatest threat to venture viability. Together, these practices reduce premature commitment to an attractive but untested idea.

A useful hypothesis is falsifiable: evidence can show that it is incorrect. “Customers will love our platform” is not sufficiently precise. “Independent restaurants with fewer than 20 employees will pay at least $80 per month to reduce weekly food-waste costs by 10%” is more testable because it identifies a customer, a benefit, a threshold, and a payment condition.

Learning Objectives

By the end of this chapter, learners should be able to:

1. Distinguish an opportunity, a problem statement, a solution concept, and a business model hypothesis.
2. Translate an opportunity into falsifiable problem, customer, solution, channel, revenue, and growth hypotheses.
3. Identify and classify assumptions by importance, uncertainty, and evidence.
4. Apply assumption mapping to select a riskiest assumption for testing.
5. Design proportionate experiments and interpret results without overstating evidence.
6. Recognize ethical, strategic, and operational limitations of early-stage validation.

1. From Opportunity Narrative to Testable Proposition

An opportunity narrative is often broad and emotionally persuasive. For example: “There should be a better way for small retailers to manage inventory.” Such a statement may identify an important domain, but it does not yet specify who experiences the problem, how frequently it occurs, what alternatives exist, or whether anyone will pay for improvement.

A testable venture proposition can be represented as a linked set of hypotheses:

• Problem hypothesis: A defined group experiences a sufficiently serious unmet need.
• Customer hypothesis: The people affected are identifiable, reachable, and capable of adopting or authorizing a solution.
• Solution hypothesis: A proposed intervention produces a meaningful improvement over current alternatives.
• Channel hypothesis: The venture can reach and serve customers through an economically feasible route.
• Revenue hypothesis: Customers or another payer will provide sufficient economic value to support the venture.
• Growth hypothesis: Adoption, retention, referrals, or distribution can expand beyond the initial customers.

These hypotheses are interdependent but should not be collapsed into one statement. A real problem does not guarantee willingness to pay. A desirable solution does not guarantee an affordable acquisition channel. Early testing is more effective when these dependencies are made explicit.

1.1 Problem versus Solution

A problem is a customer condition or outcome that creates cost, risk, frustration, delay, lost income, or another meaningful disadvantage. A solution is a proposed response to that condition. Confusing the two creates confirmation bias: founders ask whether people like the proposed product rather than whether the underlying problem is important.

A problem statement should identify:

• the affected person or organization;
• the situation in which the problem occurs;
• the current behavior or workaround;
• the consequence of the problem; and
• evidence that would indicate the problem is material.

For example, “commuters need an app” is solution-led. “Urban bicycle commuters lose time locating secure parking and often lock bicycles in unsuitable locations” is problem-led. The second statement permits investigation of existing behaviors before a product is designed.

1.2 Jobs, Pains, and Gains

The Jobs-to-be-Done perspective views customers as attempting to make progress in a particular circumstance. A “job” may be functional, emotional, or social. A university student may be trying to submit assignments reliably, reduce anxiety about deadlines, and appear organized to peers. A venture should investigate the complete job rather than only a feature request.

Pain describes the negative consequences or obstacles surrounding the job. Gain describes the desired improvement. These concepts are useful for discovery, but they are not proof of demand. Customers may describe a pain without prioritizing it, or express a desired gain without changing behavior or paying for it.

2. The Hypothesis Architecture

A practical opportunity frame can be written as follows:

For [specific customer segment] who experience [observable situation and problem], [proposed solution] will produce [measurable improvement]. We will reach them through [channel], earn revenue through [mechanism], and grow through [repeatable driver].

Each component requires its own assumptions and evidence standard.

The numerical thresholds are not universal standards. They are decision rules selected by the team. A threshold should be realistic, observable, and connected to a decision such as continuing, changing, or abandoning the proposition.

2.1 Customer Roles and the Buying System

In consumer markets, user and payer may be the same person. In organizational markets, the user, buyer, budget holder, procurement officer, technical evaluator, and beneficiary may be different. A workflow product used by nurses may be purchased by a hospital administrator and approved by an information-security team. Testing only with users can therefore produce false confidence.

A customer hypothesis should specify at least:

• user: who uses the product;
• economic buyer: who controls the budget;
• decision maker: who authorizes adoption;
• influencer or gatekeeper: who affects the decision; and
• beneficiary: who receives the outcome.

The relevant segment is not simply demographic. It should be behaviorally and contextually meaningful. “Small businesses” is usually too broad; “independent dental practices with two to five dentists that currently outsource appointment reminders” is more useful.

2.2 Channels as Hypotheses

A channel is not merely a communication medium. It includes the route by which awareness, evaluation, purchase, delivery, and support occur. Possible channels include direct sales, partnerships, marketplaces, search, communities, professional associations, and product-led acquisition.

A channel hypothesis should state the expected relationship between effort and outcome. For example: “A targeted webinar partnership with accounting firms will produce at least 15 qualified leads per event, with at least 20% agreeing to a discovery call.” This is more useful than “We will use partnerships.”

Channel testing should consider both effectiveness and economics. A channel that produces customers but requires unsustainably high labor may not be viable.

2.3 Revenue and Economic Hypotheses

Revenue hypotheses specify who pays, what triggers payment, how much is paid, and when payment occurs. Common models include subscriptions, transaction fees, licensing, usage-based pricing, advertising, commissions, and freemium conversion.

A simple contribution model can support early reasoning:

Here, Price is revenue received per customer over the selected period, and Variable cost includes costs that change with serving that customer during the same period. The expression excludes fixed costs such as salaries and rent. It is useful for comparing early customer economics, but it does not by itself establish profitability.

If customer acquisition cost is estimated, a simplified payback period can be expressed as:

Customer acquisition cost is the average acquisition expenditure per newly acquired customer, and monthly contribution is monthly revenue minus monthly variable cost. The formula assumes that contribution is positive and reasonably stable. It is not appropriate when revenue is highly seasonal, retention is unknown, or acquisition spending cannot be allocated reliably.

For example, if acquisition cost is $120 and monthly contribution is $30, the payback period is four months. This does not mean the customer is profitable after four months: retention, support costs, refunds, overhead, and the time value of money may materially alter the result.

2.4 Growth Hypotheses

Growth should be explained by a mechanism, not merely a target. Potential mechanisms include:

• repeat purchases;
• referrals;
• network effects;
• content or search accumulation;
• sales expansion within an account;
• partnerships; and
• viral or collaborative use.

A growth hypothesis must identify the behavior that produces additional adoption. “The market is large” is not a growth mechanism. “Teams invite external collaborators, and 25% of invited collaborators create their own workspace” is testable.

A basic retention measure can be defined as:

The definition must state whether new customers acquired during the period are excluded. Cohort analysis is generally preferable because it compares customers who started at a similar time. Retention is especially important for subscription ventures: acquisition without continued use may indicate curiosity rather than durable value.

3. Assumption Mapping

An assumption is a proposition treated as true for planning purposes but not yet adequately verified. Every business model contains assumptions, including assumptions about behavior, technology, regulation, costs, competition, timing, and organizational capability.

Assumption mapping makes uncertainty visible. A common two-dimensional map uses:

1. Importance or impact: How damaging would it be if the assumption were false?
2. Evidence or confidence: How much credible support currently exists?

High-impact, low-evidence assumptions deserve early attention. They are often called critical, pivotal, or riskiest assumptions. The exact label matters less than the discipline of testing the assumptions that could invalidate the venture.

Figure: A two-dimensional assumption map. High-impact assumptions with weak evidence normally receive priority for testing.

3.1 Types of Assumptions

Assumptions can be classified in several complementary ways:

Classification prevents a team from treating all uncertainty as a product-design problem. For example, a technically feasible service may still fail because customers cannot switch from a trusted incumbent.

3.2 Impact, Uncertainty, and Testability

A useful prioritization approach considers three variables:

• Impact: consequence of being wrong;
• uncertainty: weakness or absence of evidence; and
• testability: speed, cost, and ethical feasibility of obtaining evidence.

A qualitative priority score can be used when it improves discussion, but it should not create false precision. If a team assigns impact and uncertainty scores from 1 to 5, a simple screening product may be calculated as:

The score is a prioritization aid, not a probability of failure or a financial forecast. It assumes the scales are used consistently. A low-scoring assumption may still require immediate attention if it concerns safety, privacy, or legal compliance.

4. Designing Riskiest-Assumption Tests

Testing begins by converting an assumption into a decision-ready hypothesis. A strong hypothesis includes:

1. a defined population or context;
2. a predicted behavior or outcome;
3. a threshold for success;
4. a time frame; and
5. a decision rule.

Example:

Among independent tutors who currently manage bookings through messaging applications, at least 30% of qualified prospects will schedule a trial within two weeks after seeing a realistic service demonstration. If fewer than 15% schedule, the team will revisit the problem or segment before building further functionality.

The test must match the uncertainty. If the uncertainty concerns whether the problem exists, interviews and observation may be appropriate. If it concerns willingness to pay, a pricing conversation, paid pilot, deposit, or pre-order provides stronger evidence than a preference survey. If it concerns technical feasibility, a prototype or engineering spike is required.

4.1 Evidence Strength

Evidence varies in credibility. A useful hierarchy is not absolute, but it helps distinguish claims from behavior:

1. founder opinion;
2. generalized market reports;
3. stated opinions in interviews;
4. observed current behavior;
5. commitment of time or access;
6. financial commitment or payment; and
7. repeated use, retention, and referral.

Higher-level evidence is not always available at the beginning, and ethical or practical constraints may limit experimentation. Nevertheless, teams should avoid treating enthusiastic conversation as equivalent to adoption.

Interviews are particularly valuable for learning about context, language, workarounds, and decision processes. They are weak evidence for forecasting market size or confirming that a person will pay. Leading questions such as “Would you use an app that solves this?” invite politeness and speculation. Better questions ask about the last time the problem occurred, what the person did, what it cost, and what alternatives were considered.

4.2 Experiment Selection

A minimum viable product should be understood as the smallest credible vehicle for learning, not necessarily the smallest product that can be built. A concierge test may manually deliver a service behind a simple interface. A smoke test may measure interest, but it should not misrepresent a product as available or collect payment without clear disclosure.

5. Interpreting Results and Updating the Model

Validation is not the same as collecting positive comments. Teams should define what result would change their mind before conducting the test. This reduces motivated reasoning and retrospective reinterpretation.

A result can be classified as:

• confirmed enough for the next stage: evidence meets the predefined threshold;
• inconclusive: the test was too weak, poorly targeted, or affected by confounding factors;
• disconfirmed: evidence falls below the threshold or contradicts the hypothesis; or
• conditionally supported: the hypothesis appears valid only for a narrower segment or context.

Disconfirmation is productive when it prevents further investment in a weak proposition. Teams may pivot by changing the customer segment, problem, solution, channel, revenue model, or growth mechanism. A pivot should be framed as a new set of hypotheses rather than an informal change of story.

A test may fail for reasons unrelated to the core assumption. Poor messaging, an inconvenient onboarding process, a low-trust channel, or an unsuitable sample can produce a false negative. Conversely, a small enthusiastic sample can produce a false positive. Evidence should therefore be triangulated across methods and customer contexts.

6. Applied Case: ReFill Campus

The following is a fictional illustrative case.

Background

A postgraduate student team observes that students frequently purchase disposable bottled drinks on campus despite the presence of water stations. The team proposes ReFill Campus, a service combining station maps, maintenance reporting, and rewards for reusable-bottle use.

Initial Opportunity Frame

The initial statement—“Students need a convenient sustainability app”—is too broad. After preliminary observation, the team reframes the opportunity:

Students who carry reusable bottles often cannot quickly identify functioning, accessible refill stations between classes. A mobile service that shows reliable station availability and rewards verified refills may increase refill behavior.

The frame exposes several assumptions. Students may not consider the inconvenience important; station data may become outdated; rewards may not change behavior; universities may be the payer; and the service may be difficult to distribute outside one campus.

Assumption Map

The team lists and ranks assumptions:

The riskiest assumption is not that an app can be coded. It is that the problem is sufficiently important to change behavior and support an institutional or consumer payment model.

Approach

The team conducts twelve contextual interviews across two campuses, observes refill decisions during class-change periods, and asks participants to recount their most recent attempt to refill a bottle. It finds that most students do not actively search for stations; they refill only when a station is directly visible. However, students living in residence halls report a different problem: stations are frequently empty or poorly maintained.

The team then creates a low-fidelity map and manually verifies station status twice daily for one residence hall. Participants receive access to the map for two weeks. The team measures station checks, refill attempts, reported failures, repeat use, and willingness to pay. It also interviews facilities managers about maintenance data and procurement constraints.

Outcome

The evidence weakens the original student-subscription hypothesis. Students appreciate accurate information but rarely pay for it. The residence-hall group uses the service more frequently, while facilities managers show interest in aggregated failure reports. The team therefore changes the customer and revenue hypotheses: the university, rather than the individual student, may pay for a maintenance and sustainability dashboard, while students remain end users.

This is not proof of a viable business. The revised concept still requires testing of procurement timelines, data integration, institutional budgets, measurable sustainability outcomes, and the cost of maintaining accurate information. Nevertheless, assumption mapping prevented the team from spending heavily on rewards and a consumer subscription before understanding the problem and payer.

Learning Points

1. The observable problem differed across student segments.
2. Users and economic buyers were not the same.
3. A technical prototype tested less important uncertainty than behavior and payment.
4. Manual delivery was useful for learning about value before automation.
5. Evidence supported a narrower opportunity rather than the original broad claim.

7. Managerial and Professional Implications

Opportunity framing is a governance practice as well as a discovery technique. It creates a shared language for founders, investors, employees, partners, and educators. Explicit hypotheses make resource allocation more defensible because a team can explain what it is trying to learn and why.

Managers should maintain an assumption register containing the assumption, owner, evidence, date, test, result, and decision. The register should be revised as market conditions change. Assumptions about regulation, platform access, supplier costs, and competitive behavior may become riskier even after customer demand is established.

Teams should also distinguish evidence from interpretation. A statement such as “eight of ten interviewees liked the prototype” is evidence of expressed approval in a particular sample. “The market wants the product” is an interpretation that requires stronger support.

8. Ethics, Governance, and Responsible Experimentation

Customer discovery involves real people, personal data, and potentially unequal power relationships. Ethical practice requires informed participation, honest representation of prototypes, protection of sensitive information, and respect for the right to decline. Teams should not exploit urgency, financial vulnerability, or employment dependence to obtain commitments.

Experiments involving health, finance, children, employment, safety, or protected personal data may require specialist review and formal compliance processes. A landing page that tests demand should not claim that a service is available when it is not. Payment experiments must state refund, cancellation, and data-use conditions clearly.

AI can support transcription, clustering of interview themes, drafting of hypotheses, and simulation of alternative messages. It should not replace direct customer evidence or be treated as a representative customer. Automated analysis may reproduce bias, omit minority perspectives, or expose confidential information. Human review, data minimization, access controls, and transparent documentation are essential.

9. Limitations and Common Errors

Several recurring errors undermine assumption mapping:

• Feature fixation: testing preferences for features before establishing the problem.
• Vague segmentation: treating a large demographic category as a customer segment.
• Vanity metrics: counting views, downloads, or sign-ups without activation or retention.
• Leading interviews: prompting agreement rather than investigating behavior.
• Premature scaling: investing in acquisition before retention and economics are credible.
• False precision: assigning numerical probabilities unsupported by evidence.
• Ignoring non-users: studying only enthusiastic early adopters.
• Failure to update: preserving the original narrative after contradictory results.
• Confusing feasibility with desirability: assuming that a buildable product is valuable.

No experiment eliminates uncertainty. The objective is to reduce the most consequential uncertainty at an acceptable cost and to make a better next decision.

Best-Practice Workflow

A disciplined workflow is:

1. State the opportunity without prescribing a solution.
2. Define the narrowest plausible customer segment and relevant roles.
3. Write separate problem, customer, solution, channel, revenue, and growth hypotheses.
4. List assumptions behind each hypothesis.
5. Rate impact and evidence quality; identify high-impact, low-evidence assumptions.
6. Define a falsifiable prediction, threshold, time frame, and decision rule.
7. Select the smallest ethical experiment that can produce meaningful evidence.
8. Observe behavior and commitments, not only opinions.
9. Record results and alternative explanations.
10. Continue, modify, pause, or abandon the proposition based on evidence.

Chapter Summary

Opportunity framing translates an entrepreneurial idea into a connected set of hypotheses about a problem, customer, solution, channel, revenue model, and growth mechanism. Good hypotheses are specific, observable, and falsifiable. They distinguish the underlying customer problem from the proposed solution and recognize that users, buyers, decision makers, and beneficiaries may be different people.

Assumption mapping exposes the beliefs on which a venture depends. High-impact assumptions supported by weak evidence are normally the riskiest and should be tested before lower-impact technical or cosmetic decisions. Tests should match the uncertainty: interviews and observation help investigate context and behavior; prototypes test comprehension and use; paid pilots and deposits provide stronger evidence about willingness to pay; and cohort observation is needed for retention.

Evidence must be interpreted cautiously. Positive opinions are weaker than repeated use or payment, and a failed experiment may reflect poor design rather than a false core assumption. Ethical experimentation requires honesty, privacy protection, appropriate consent, and special care in regulated or vulnerable contexts. The purpose of discovery is not to prove a preferred idea correct, but to improve decisions by learning what must be true for the opportunity to become viable.

Key Terms

Opportunity: A perceived possibility to create value by addressing a need or inefficiency.

Opportunity framing: Expressing an opportunity as connected, testable business hypotheses.

Hypothesis: A proposition treated as provisional and subject to evidence.

Falsifiability: The property of being capable of being contradicted by evidence.

Problem hypothesis: A claim that a defined customer group experiences a meaningful difficulty.

Customer segment: A relatively specific group sharing relevant needs, context, or behavior.

Economic buyer: The person or organization controlling the budget for a purchase.

Solution hypothesis: A claim that a proposed intervention will improve a customer outcome.

Channel hypothesis: A claim about how customers can be reached, converted, served, or supported.

Revenue hypothesis: A claim about who pays, what they pay for, and under what terms.

Growth hypothesis: A claim about the mechanism by which adoption expands.

Assumption: An unverified belief required for a plan or model to work.

Assumption map: A framework that prioritizes assumptions according to impact and evidence.

Riskiest assumption: A high-consequence belief supported by insufficient evidence.

Minimum viable product: The smallest credible product or service arrangement designed to generate learning.

Concierge test: A manually delivered service used to test value before automation.

Validation: Evidence-based support sufficient to justify the next stage of investigation or investment.

Retention: Continued customer activity or use over a defined period.

Pivot: A substantive change to a venture hypothesis based on learning.

Vanity metric: A measure that appears positive but does not reliably indicate venture progress.

Review Questions

Short Answer Questions

1. What makes an entrepreneurial hypothesis falsifiable?
2. Why should a problem be investigated before a solution is built?
3. Distinguish a user, an economic buyer, and a decision maker.
4. What two dimensions are commonly used in an assumption map?
5. Why is repeated use generally stronger evidence than stated interest?

Long Answer Questions

1. Explain how a broad opportunity narrative can be translated into problem, customer, solution, channel, revenue, and growth hypotheses.
2. Discuss the strengths and limitations of interviews, prototypes, paid pilots, and cohort tests as customer-discovery methods.
3. Explain how impact, uncertainty, and testability should influence assumption prioritization.
4. Analyze why a technically feasible product may still fail as a business opportunity.
5. Discuss the ethical requirements for experimentation involving personal data, vulnerable customers, or regulated sectors.

Analytical / Application Questions

1. A team proposes a subscription service for independent restaurants that predicts ingredient demand. Write one falsifiable hypothesis for each of the six opportunity elements and identify the riskiest assumption.
2. A landing page receives 1,000 visits, 80 email sign-ups, and no paid deposits. What can reasonably be inferred, what cannot be inferred, and what experiment should follow?
3. Using the ReFill Campus case, explain why changing the payer from students to universities represents a change in the business model hypothesis rather than proof that the venture is validated.

Multiple Choice Questions with Answers

Question 1

Which statement is the most falsifiable hypothesis?

A. Customers will love our platform.

B. The market for productivity tools is large.

C. Independent retailers with one to three stores will pay $75 per month for a stock-reconciliation service.

D. Our product will transform retail.

Correct Answer: C. Independent retailers with one to three stores will pay $75 per month for a stock-reconciliation service.

Question 2

Which assumption is normally the highest priority for early testing?

A. The logo color is attractive.

B. The software can display a dashboard.

C. The target customers have a serious problem and will pay for improvement.

D. The product can later add a dark mode.

Correct Answer: C. The target customers have a serious problem and will pay for improvement.

Question 3

Who is the economic buyer?

A. The person who experiences the problem in every case

B. The person or organization controlling the purchase budget

C. The person who first hears about the product

D. The person who designs the user interface

Correct Answer: B. The person or organization controlling the purchase budget

Question 4

Which form of evidence is generally stronger than a positive interview opinion when testing willingness to pay?

A. A larger advertising slogan

B. A paid pilot or deposit

C. A founder’s intuition

D. A competitor’s website visit

Correct Answer: B. A paid pilot or deposit

Question 5

What is the main purpose of a concierge test?

A. To automate the entire service immediately

B. To test value through manual delivery before extensive automation

C. To estimate the total population of a market

D. To replace all customer interviews

Correct Answer: B. To test value through manual delivery before extensive automation

Question 6

If customer acquisition cost is $120 and monthly contribution is $30, what is the simplified payback period?

A. 2 months

B. 3 months

C. 4 months

D. 6 months

Correct Answer: C. 4 months

Question 7

Which statement best describes a growth hypothesis?

A. The market is very large.

B. The product has many possible features.

C. Each retained customer will generate at least one qualified referral every three months.

D. The founding team is highly motivated.

Correct Answer: C. Each retained customer will generate at least one qualified referral every three months.

Question 8

Why are vanity metrics potentially misleading?

A. They are always inaccurate mathematically.

B. They may look positive without demonstrating activation, retention, or economic value.

C. They can only be collected by large firms.

D. They cannot be displayed in a dashboard.

Correct Answer: B. They may look positive without demonstrating activation, retention, or economic value.

Question 9

In the ReFill Campus case, what important discovery changed the business model?

A. The map could not be built.

B. Students had no interest in sustainability.

C. Universities appeared more plausible as payers than individual students.

D. Rewards were proven to increase retention.

Correct Answer: C. Universities appeared more plausible as payers than individual students.

Question 10

Which practice is most consistent with responsible experimentation?

A. Hiding that a prototype is not yet available

B. Collecting unnecessary personal data to increase future options

C. Defining decision rules in advance and clearly explaining the experiment

D. Treating artificial-intelligence output as a substitute for customer evidence

Correct Answer: C. Defining decision rules in advance and clearly explaining the experiment