Skip to main content

Module 2 · Customer Discovery & Problem Validation

2.2 Problem Validation & Customer Understanding

Skills Track

Whom should I understand first, and how do I turn conversations into an evidence-informed problem statement?

So far, we have learned to approach people's experiences with curiosity, ask questions that do not push toward a preferred answer, and respect the ethical boundaries of inquiry. That is an important beginning, but it creates a new challenge. Suppose you now have several conversations filled with detailed stories. Some people describe the same frustration; others disagree about what matters. A few have tried workarounds, while others seem content with the current process. How do you make sense of these differences without forcing everyone into one convenient category? This section develops the skills of defining whom to understand, recognizing meaningful patterns, and describing problems carefully enough that other people can evaluate the reasoning behind your conclusions.

2.2.1 Defining Whom to Understand

2.2.1.1 The person using a service may not be the person choosing or paying for it

When someone says “the customer,” we often picture one person making one decision. In many real settings, a problem involves several people with different responsibilities and interests. The person who uses a product may not purchase it. The person who pays may not choose which product will be bought. A manager who approves a change may not experience the everyday inconvenience that prompted the proposal. If we listen to only one person, we may develop an excellent understanding of one perspective and a misleading understanding of the entire situation.

Consider the purchase of scheduling software for a small repair company. Technicians use the schedule throughout the day. An office coordinator may create appointments and send confirmations. Customers care about when the service will arrive. The business owner may approve the budget, and an accountant may review the cost. Each person has a legitimate concern, but the concerns differ. Technicians may value accurate job locations and enough travel time. Customers may want reliable arrival windows. The owner may worry about the subscription cost and whether the software will be adopted. A system that pleases one group while making another group's work impossible may not create sustainable value.

This gives us several useful roles. A user directly interacts with a product or service. A buyer or purchaser acquires it. An economic buyer has authority over the expenditure or budget approval, especially in organizational purchases. A decision-maker selects or authorizes an action, sometimes with input from others. A stakeholder is someone affected by the decision even if they never purchase or directly use the offering. These roles can overlap, and their meaning varies by setting. The point of distinguishing them is not to memorize business vocabulary for its own sake; it is to avoid assuming that the loudest or most visible person represents everyone whose experience matters.

The same framework is useful outside business. A college may purchase technology for students to use, but administrators, instructors, support personnel, and students all have different perspectives on accessibility and effectiveness. A government agency may design a public-service form used by residents but evaluated by staff who enforce legal requirements. A healthcare organization may acquire equipment used by clinicians but directly affecting patients. In each case, discovery becomes more responsible when the investigator maps the people who use, authorize, support, and are affected by the decision before announcing what “the customer” needs.

2.2.1.2 Users, early adopters, and people most affected

Sometimes an early investigation should focus on people who experience a difficulty frequently or intensely. In a business context, such individuals may become early adopters—people who have an unusually strong reason to try a new solution before it is widely established. They are not necessarily enthusiastic about technology or novelty. A homeowner who has repeatedly missed critical repair appointments might be open to an unfamiliar scheduling service because the existing options have caused substantial inconvenience. A small business owner who spends hours each week reconciling invoices may be willing to consider a new process simply because the present workload is unsustainable.

A useful question is therefore not just “Who could theoretically benefit?” but “Whose present circumstances make the problem especially consequential?” People who already invest significant time, effort, money, or social support in coping with a difficulty may provide particularly informative accounts of what happens and why. Their workarounds can show what they value and which constraints a possible improvement would need to respect. But early adopters are not necessarily representative of everyone who might eventually use a service. Their urgency and willingness to tolerate imperfections can differ considerably from the preferences of a broader population.

An investigator must also distinguish people who experience the problem from people who have power to approve a solution. Imagine a school purchasing a digital platform to help students organize assignments. Students may be the primary users, educators may configure assignments, and the institution may pay. If an entrepreneur interviews only students, they could misunderstand purchasing requirements. If they interview only administrators, they could overlook the actual user experience. A responsible professional would need to consider both forms of evidence while recognizing that the purchasing process is a separate question from whether the proposed solution works for students.

Research on the lean startup framework describes customer development as an effort to examine hypotheses about customer groups, value propositions, and how a venture might reach people (Shepherd & Gruber, 2021). That framework is especially useful when treated as an invitation to identify uncertainties, rather than as a formula for selecting a profitable segment without further inquiry. In Module 2, our purpose is still to understand problems. We will examine purchasing decisions and business models more fully in Module 3. For now, the learning task is to find out whose experience needs to be represented before a problem can be described accurately.

2.2.1.3 A stakeholder map as a thinking tool

One simple way to avoid overlooking a perspective is to create a stakeholder map. Rather than beginning with an organization chart, start with the problem as people experience it. Place the affected person or process in the center and identify who participates at different moments. Who experiences the difficulty? Who helps someone work around it? Who controls important resources? Who makes the rules? Who bears costs when the process fails? A map cannot tell you what any of those people believe, but it can reveal who has not yet been heard.

For Elena's clinic example, the map could include visitors preparing for appointments, reception staff, clinicians, administrative managers, and people supporting visitors with transportation or documents. Elena should not infer that every group has equal authority or access to information. Nor should she invite unapproved interviews with patients merely because the map suggests their experiences matter. The tool helps her see the structure of the situation so she can decide, with appropriate oversight, what kind of information is ethically and practically available.

The surprising value of a stakeholder map is that it can reveal conflicting definitions of success. A manager may value shorter check-in times; a patient may value clear guidance that preserves dignity; a clinician may value accurate records and safe preparation; a receptionist may value procedures that are realistic during peak periods. These goals are not automatically incompatible, but they cannot be reconciled if the team pretends everyone wants exactly the same thing. Good discovery identifies the tensions before they become costly surprises.

2.2.2 Customer Segmentation and the Minimum Viable Customer Segment

2.2.2.1 Why “everyone” is usually too broad a group to understand

Imagine that an organization says its intended audience is “all college students.” That sounds inclusive, but it may hide important differences. Some students attend full-time during the day. Others take evening classes while working. Some are new to college and unfamiliar with academic procedures. Others are returning adults with considerable workplace experience. Some study entirely online, and some need specific accommodations to access materials. A service designed around one imagined “typical student” may function poorly for many of the people it claims to serve.

Segmentation means grouping people according to characteristics that are meaningful for the problem being investigated. Age or income can sometimes matter, but those categories may tell us less than behavior, circumstances, needs, frequency of a difficulty, or access to alternatives. For questions about advising access, an evening student's work schedule may be more relevant than their age. For questions about using a digital service, access to a compatible device and reliable internet may matter more than a broad label such as “young adult.” Good segmentation begins with the problem and asks which differences actually change the experience or required response.

The Jobs to Be Done perspective reinforces this approach by asking which outcome a person seeks and what conditions shape their choices (Christensen et al., 2016). Two people of very different ages may share the same pressing need to arrange appointments outside conventional work hours. Conversely, two people of similar age and income may use the same service for completely different reasons. Segmentation is therefore not a license to stereotype. It is a disciplined attempt to identify useful similarities and differences so that research and design can become more relevant.

This matters in professional practice. A library considering evening access might distinguish shift workers, online learners needing equipment, and parents attending family activities because their patterns of use differ. A healthcare organization might examine access barriers by language needs, appointment type, or transportation constraints while honoring applicable privacy and nondiscrimination requirements. Segmenting responsibly means considering how to serve different people more effectively, not simply selecting those who are easiest or most profitable to help. In business, where resource choices also matter, such distinctions eventually influence how an offering is designed and sustained.

2.2.2.2 Choosing a narrow first learning group without shrinking the mission

A beginner investigating an opportunity often faces a choice between an enormous audience and a tiny set of convenient acquaintances. The first is difficult to study meaningfully; the second may provide a distorted picture. A useful middle step is to define a Minimum Viable Customer Segment (MVCS): a narrow, provisional group whose members share circumstances relevant enough to the problem that the investigator can learn from them coherently. In this textbook, MVCS is a practical planning label for the initial learning scope, not a claim that the group is large enough to support a viable business.

Suppose a community college wants to understand why some students struggle to use advising services. An initial group might be students enrolled in evening classes who also work rotating shifts. That group is narrow enough to explore a recognizable constraint, but it is not defined merely by an arbitrary demographic label. Another investigation might focus on students who recently transferred from another institution and are trying to interpret degree requirements. The choice of learning group should follow the question, not the convenience of recruiting people who are already within reach.

Narrowing the first group can make contradictions easier to interpret. If participants have widely different circumstances, a researcher may hear ten different problems and struggle to determine what they have in common. Focusing on a shared situation gives the investigator a clearer picture of which obstacles vary and which recur. That does not justify ignoring other groups forever. It means being transparent about the question currently being studied and expanding the inquiry when the evidence suggests important differences. The method is similar to examining one neighborhood intersection carefully before claiming to understand traffic throughout an entire city.

Blank (2013) describes customer development as a process of learning about customers before committing to a business model. The practical implication for this introductory module is that early scope should support learning rather than create an illusion of market certainty. If you define a group, explain why it is relevant, who may be excluded, and what you still do not know about its size or needs. In the next module, you will encounter broader market analysis and business model choices. For now, a small, well-explained learning group is more valuable than a grand claim that “everyone” will benefit.

2.2.2.3 A segment is a hypothesis about similarities

There is a danger in segmentation that deserves particular attention: once we name a group, we may start treating its members as though they are all alike. If an investigator labels people “busy parents,” they may assume all have the same scheduling preferences. If a team describes “small business owners,” it may overlook differences between a solo tradesperson, a family restaurant, and a growing repair company. A useful segment is not a box into which every member must fit. It is a hypothesis that certain shared circumstances may explain relevant behavior.

Consider three hypothetical people using the same public transportation route. One is a student whose classes end late. One is a caregiver who needs to reach an appointment. One is a worker whose shift begins before sunrise. They share transportation use, but the consequences of a missed connection may differ. The investigator might discover that reliability is important to all three and that each defines an acceptable schedule differently. Instead of immediately merging them into one “transit customer” segment, the researcher can investigate where common needs end and distinct constraints begin.

Segmentation should remain revisable. New evidence may show that a category is too broad, too narrow, or organized around the wrong variable. It may also reveal that a group originally thought marginal faces the most severe barriers. Intellectual honesty requires changing the categories when the data no longer support them. This is part of entrepreneurial learning: we use temporary structures to make complexity understandable, but we must not confuse those structures with the full reality of people's lives.

2.2.3 Understanding and Interpreting Interviews

2.2.3.1 Before the conversation: decide what you are trying to learn

A meaningful interview begins before the first question. The investigator identifies the main uncertainty, selects people whose experiences are relevant, and decides what evidence could help distinguish among competing explanations. A broad objective such as “learn what students think about advising” may lead to a friendly conversation but provide little guidance for interpretation. A narrower objective—“understand what happened when working students tried to get advising appointments during the past month”—creates a clearer path to useful stories. Good preparation does not eliminate surprises; it gives the interviewer a way to recognize why a surprising answer matters.

Preparation also means knowing what is outside the scope of the interview. An investigator examining appointment access may not need to ask about medical diagnoses, detailed finances, or family histories. If the question concerns a workplace process, the interviewer should not collect confidential client information or proprietary records without authorization. Setting boundaries in advance protects people and often makes the research more focused. It also helps the interviewer explain the purpose honestly and avoid the temptation to collect every interesting detail simply because it was offered.

A semi-structured interview balances planned questions with open follow-up. It might begin with a recent event, examine the steps the person took, explore what made the experience difficult or successful, and end by asking what else the investigator should understand. The same general themes appear across conversations, but the exact order can vary. Brinkmann and Kvale (2015) emphasize that interviewing involves both a method and an interpretive relationship: the researcher shapes the conversation while trying to understand the participant's account. That is why good questions require careful listening, not simply a technically correct list.

Imagine interviewing someone about how they chose a childcare arrangement. Asking for a recent example may reveal trade-offs among cost, location, schedule, trust, and availability. The person may emphasize an issue that the interviewer never anticipated, such as a changing work schedule or the need for care during a school break. An inflexible script would steer the conversation back to predetermined categories. A prepared but open interviewer would recognize that the new detail may be essential to understanding the decision. This is the difference between filling blanks and discovering something.

2.2.3.2 During the conversation: separate the person's words from your interpretation

When we listen, we naturally translate what people say into familiar categories. Someone describes a confusing website and we write “bad navigation.” Someone describes expensive transportation and we write “price-sensitive customer.” These summaries may be plausible, but they are not identical to the original account. If we replace people's language too quickly, we can lose the specific circumstances that gave the story meaning. A stronger note-taking habit distinguishes the person's description, the observed behavior if any, and the interviewer's emerging interpretation.

Consider three notes from fictional conversations about advising access:

Illustrative participant accountObservable or reported eventPossible interpretation—not established fact
“I called twice during my break and couldn't get through.”Two attempts made within a limited time windowPhone availability may conflict with work schedules
“I wasn't sure which office to ask.”The person reported uncertainty about where to beginService information may be difficult to navigate
“My friend showed me how to do it online.”The person relied on peer assistanceInformal support may compensate for unclear instructions

The table makes something important visible: interpretations are not evidence simply because they appear next to evidence. We still need to ask whether other explanations fit the facts. A person who called twice may have chosen an unusually busy moment; a confusing office name may be one part of a larger information problem; peer assistance may be convenient rather than necessary. Good notes preserve enough detail to revisit those possibilities later instead of locking the first interpretation into place.

Listening also means attending to exceptions and positive experiences. If someone navigated the process easily, ask what made it work. Perhaps a support person explained the steps clearly, the individual had previous experience, or a particular resource was accessible. These accounts may reveal features worth preserving or expanding. A discovery process that records only complaints can produce a distorted picture just as easily as one that records only compliments. The aim is to understand the process, not to build a case against it.

2.2.3.3 After the conversation: look for patterns without erasing differences

After several conversations, the investigator can begin organizing the information into themes—patterns of meaning that help explain the experience being studied. A theme might involve limited appointment availability, confusing instructions, lack of transportation, or reliance on informal help. Themes are more informative when they include examples, context, and evidence of variation. They are less informative when they become vague categories such as “communication issues” that could mean almost anything. Strong analysis asks what the theme actually describes and how it connects to the research question.

Braun and Clarke (2006) describe thematic analysis as a flexible approach to identifying and interpreting patterns in qualitative information. Their contribution should not be mistaken for a mechanical rule that counting words automatically reveals truth. A repeated phrase may reflect the interview guide, a shared cultural expression, or a meaningful recurring problem. The investigator must interpret the context and explain the basis for the theme. Careful qualitative analysis involves reviewing the material, developing tentative categories, checking them against the original accounts, refining the categories, and communicating the findings with appropriate limits.

Imagine that six fictional interviewees describe difficulties with a campus appointment system. Four mention a problem finding available times, three mention unclear instructions, and two mention transportation. The categories overlap; one person may report more than one issue. These counts could help an investigator organize what was heard, but they do not establish prevalence across the college. Nor do they mean appointment times are necessarily the “biggest” problem in the institution. The two transportation accounts might describe more serious consequences than the four scheduling frustrations. Frequency and importance are related but not identical.

A thoughtful summary might say, “Among the six illustrative accounts, difficulty finding workable appointment times appeared repeatedly, often alongside other constraints. Two accounts described transportation as a major factor. Because the group was small and selected for relevance rather than representativeness, further investigation would be needed before drawing conclusions about the wider student body.” This may sound less dramatic than “We've validated the problem.” It is more academically credible and more useful to someone who must make a responsible decision.

2.2.3.4 Contradictions can reveal the real boundaries of a problem

An inexperienced investigator often worries when people disagree. If four participants say a process is difficult and two say it works well, the results may feel confusing. But disagreement can provide precisely the information needed to understand which circumstances make the process succeed or fail. Perhaps the successful participants have reliable internet, previous experience, or appointments at different times. Perhaps the struggling participants face a particular combination of constraints. If you collapse all six answers into an average level of satisfaction, you may miss the mechanism behind the variation.

This is one reason researchers examine negative cases or disconfirming cases—accounts that do not fit an emerging explanation. Suppose Elena develops the tentative claim that unclear reminders cause missed documents. An account from someone who received a clear reminder yet still lacked documents suggests that reminder clarity is not a complete explanation. It does not necessarily disprove that reminders matter for other people. It tells us that the statement needs to be qualified or divided into separate problems. A person who welcomes contradictory evidence may reach a more precise and useful understanding than someone who tries to eliminate it.

The skill also helps students in academic life. When reading a case study, a source may describe a trend while another presents an exception. The responsible response is not always to choose which source sounds more persuasive. It is to ask whether the sources studied different populations, contexts, or moments in time. Entrepreneurial learning and academic reasoning meet here: both improve when we treat differences as clues to investigate rather than inconveniences to hide.

Entrepreneur-in-Action

Bank of America and IDEO: Understanding a Savings Behavior

A documented example of design research comes from work involving IDEO and Bank of America on the Keep the Change savings service. Rather than beginning exclusively with a conventional list of banking features, the work drew on an understanding of consumer behavior and how people manage money. IDEO has discussed how behavioral observations and analytics helped inform the customer groups explored during the project (IDEO, 2008). The resulting service became associated with automatically rounding eligible debit-card purchases and transferring the difference into savings. The example is useful because it shows that a financial product can be shaped around an existing behavior rather than only an abstract claim about what customers should do.

The case is sometimes simplified into the inspirational story that designers saw someone round up in a checkbook and instantly invented a product. The more defensible lesson from IDEO's own discussion is that qualitative observation and quantitative selection of customer groups contributed to understanding a pattern of behavior. That distinction matters: a memorable anecdote can make innovation seem like a lightning strike, while actual research often involves choosing whom to study, noticing what people do, and working with an organization capable of implementing a service. The exact contribution of each research step to commercial success cannot be inferred from the case alone.

Consider the larger insight for a student. If people want to save money but struggle to set aside a fixed amount each month, the problem may not be that they lack a savings goal. It may be that the current method does not fit their ordinary routines. An educator might ask how students organize deadlines; a manager might ask how employees record small but recurring tasks; an entrepreneur might ask how customers complete repetitive transactions. In each setting, understanding an existing behavior can reveal possibilities that a simple question such as “Would you like a better product?” would never expose.

See all 24 entrepreneurship case studies →

2.2.4 Customer Profiles and Personas

2.2.4.1 A persona is a model, not a human being

Once an investigator has listened to several people, there is often a desire to make the findings easier to remember. One common tool is the persona, a short, evidence-informed description of a particular type of user or customer. A persona might summarize goals, constraints, typical circumstances, and relevant behaviors for a group. It can help teams avoid designing around themselves, especially when the people who will use a service are not present during every decision. But a persona also creates a risk: a fictional profile can begin to feel like a real person whose needs have already been established.

Imagine a team developing a persona named “Jordan, the Working Student.” The profile says Jordan is 29, takes evening classes, has two children, works in retail, prefers texting, and struggles with daytime advising. This may be a vivid character, but unless the details are supported by actual research, the team may simply be inventing a story that fits its assumptions. A more defensible profile could describe a learning segment instead: students who attend evening classes and report difficulty reaching advising during standard office hours. The description can include documented behaviors and constraints without pretending that everyone in the group has the same family structure or preferred communication channel.

The value of a persona lies in making relevant evidence portable. A designer can ask, “Would the process work for the student who has a 15-minute break between shifts?” A supervisor can ask, “How would this scheduling policy affect a technician who spends much of the day traveling between jobs?” These questions can expose practical problems during planning. Yet the persona should remain a model to be revised when new evidence appears. It is not a substitute for speaking with real people, and it should never be used as an excuse to ignore participants whose circumstances differ.

2.2.4.2 Build profiles from needs, behaviors, and constraints

A useful introductory customer profile might contain six fields: the relevant situation, the person's goal, the difficulty encountered, the current workaround, the consequences of the difficulty, and the remaining uncertainties. These fields emphasize features that can influence a solution. They also help avoid irrelevant decorative details. Knowing someone's favorite beverage or assigning a stock photograph can make a persona look complete while contributing nothing to the problem being investigated. A stronger profile is compact enough to guide a decision and transparent enough to show what is still unknown.

Consider this fictional, illustrative profile for a student-service improvement project:

FieldEvidence-informed working description
SituationEnrolled in evening courses; works shifts that vary week to week
GoalGet accurate advising guidance before a registration deadline
Reported difficultyAvailable advising hours overlap with paid work
Reported workaroundSends email, asks classmates, or delays the decision
Consequence to investigateUncertainty and possible delay in selecting appropriate classes
Still unknownHow common the difficulty is, which alternatives help, and which changes would be feasible

This profile does not claim to describe every working student. It presents a coherent hypothesis about a group whose experiences deserve investigation. It also distinguishes reported difficulty from potential consequence. The latter must be supported before it can be stated confidently as an observed outcome. That small distinction can prevent a team from accidentally turning its own interpretation into a research finding.

Personas can also include multiple perspectives. In a business that sells equipment to organizations, the end user may want ease of operation, a purchasing manager may value reliability and cost, and a supervisor may be concerned about training and compliance. One persona cannot honestly represent all three just because they interact with the same product. A stakeholder map identifies the roles; research-based profiles help explain relevant differences within those roles. Together, they give a more realistic view of a problem than a single imaginary “ideal customer.”

2.2.4.3 The danger of using stereotypes as shortcuts

When a persona is built around assumptions about age, gender, income, language, disability, ethnicity, or family structure, it can quietly reproduce stereotypes. A team might claim that older adults dislike technology or that younger students prefer mobile apps. Neither statement reliably describes every individual in the group, and either can lead to poor service design. A responsible profile distinguishes characteristics that actually affect the process from broad labels that merely feel familiar. If access to reliable internet matters, investigate internet access. If a service requires English-language instructions, examine language accessibility. Do not treat identity categories as substitutes for research into the conditions that matter.

This does not mean demographic information is always irrelevant. It can be essential when studying unequal access, identifying service barriers, or evaluating whether a sample omits important groups. The difference is whether the information is used carefully and for a justified purpose. A healthcare organization examining transportation access may need to consider how disability and geographic location affect options. A college assessing support services may need to investigate whether students with different schedules face different barriers. The ethical question is whether those distinctions help people receive better service, rather than simply giving designers a convenient stereotype to design around.

An intellectually honest persona also includes its own level of confidence. A short note such as based on four exploratory interviews; not representative of all evening students can change how a team uses the tool. It encourages the team to treat the profile as a provisional learning aid. That is consistent with the broader argument in this chapter: the purpose of customer understanding is not to create an impressive-looking file. It is to become better informed about actual people, and to know when the file no longer reflects what those people are telling us.

Venture TrackVT-2B · Optional

Companion to 2.2.4 Customer Profiles and Personas

Build an Early-Adopter Snapshot

Suppose Priya wants to offer after-hours bookkeeping support to small home-service businesses. An appealing first customer description might be “every small business owner who dislikes paperwork.” That's too vague to guide a useful investigation. A stronger early-adopter snapshot identifies a specific situation: an owner-operator who manages service calls during the day, has recurring bookkeeping delays, and currently spends personal evenings reconciling invoices. This description is a hypothesis to investigate, not a portrait of a market already proven to exist.

Startup questionWorking hypothesis to investigate
Who directly experiences the difficulty?Owner who prepares invoices and records after service calls
Who authorizes payment?Often the same owner in a solo operation; may differ in a larger company
What recent behavior would be relevant?Repeated late-night reconciliation, delayed invoices, or a paid workaround
Where might relevant people be reached ethically?Local business networking groups, introductions, or voluntary industry communities
What would challenge the hypothesis?Owners already have satisfactory support or do not view the delay as consequential

Priya should not decide that everyone in this category needs outsourced bookkeeping. She would use the snapshot to seek people with the relevant experiences and listen for variation. If owners consistently describe difficulty but have no realistic budget, the problem and the revenue opportunity still need to be evaluated separately.

What this teaches

Translate an evidence-informed segment into a focused, realistic recruitment strategy while distinguishing the user from the economic buyer.

2.2.5 Writing an Evidence-Informed Problem Statement

2.2.5.1 A good problem statement makes the experience visible

A problem statement is a concise description of a difficulty, the people affected, the situation in which it occurs, and the consequences that make it matter. It helps others understand what is being investigated without requiring them to accept a preferred solution. A weak statement might read, “Students need a better advising app.” That sentence sounds like a problem, but it actually prescribes a product. Another weak statement might say, “Advising is broken.” That communicates frustration but gives little information about who experiences which difficulty or when it occurs. Neither statement helps a thoughtful reader determine what evidence would be relevant.

A stronger working statement might be: “Some students enrolled in evening courses report difficulty obtaining advising assistance before registration deadlines when available appointment hours overlap with their work schedules, leading them to seek informal guidance or delay decisions.” This statement names a group, difficulty, context, and reported response. It deliberately uses the word some rather than suggesting the issue affects everyone. It avoids claiming that an app, extra staff, or longer office hours is already the correct response. A well-formed problem statement leaves room for inquiry instead of closing the discussion with a preferred intervention.

For beginning learners, a practical structure is: [A defined group] experiences [a specific difficulty] when [a context or triggering situation occurs], with [a documented or carefully qualified consequence]. The words in the brackets are prompts, not a rigid template that every human experience must fit. Some problems are complex enough to require more than one sentence. Others involve several stakeholders or processes. The purpose is to force clarity about the relationship between people, conditions, and consequences before moving toward design.

This writing discipline matters in almost every career. An employee who says “Our team needs new software” may get resistance from a manager concerned about cost. An employee who can explain that staff repeatedly reenter the same information into two approved systems, document the time involved, and identify where errors occur has a stronger basis for requesting a review. The difference is not clever rhetoric. It is the ability to define the problem in terms that other people can examine. That is both an entrepreneurial and an academic capability.

2.2.5.2 What belongs in the statement—and what should stay out

An evidence-informed problem statement includes only claims that the current information can reasonably support. If a student has interviewed four people, the statement should not say “most customers.” If a person reports feeling frustrated, the investigator can describe the report but should not claim to know the person's internal experience beyond what was shared. If a process appears to cause delays, the writer should identify whether the relationship was observed or inferred. The goal is not to make every sentence timid. It is to make the strength of the language match the strength of the evidence.

The statement should also avoid silently incorporating a solution. “Drivers need an AI scheduling assistant” tells us what a designer wants to create. “Some drivers report spending unpaid time coordinating last-minute schedule changes through multiple messages” describes a difficulty that could be addressed in several ways. A better communication policy, revised staffing procedure, shared calendar, or technology tool might eventually help. At the discovery stage, the solution remains an open question. Protecting that openness gives the investigator a chance to compare alternatives in Module 3 and beyond.

An especially useful addition is a brief evidence note: what information supports the statement, how it was gathered, and what remains unknown. For example: Based on five exploratory interviews with evening students; appointment availability and work schedules were recurring concerns; the prevalence of this difficulty across the college is unknown. This note turns a polished sentence into an honest representation of what has been learned. It also gives instructors, supervisors, or potential partners the ability to challenge the reasoning without dismissing the experience of the people who spoke.

In the clinic scenario, Elena might revise her initial statement from “Patients do not understand appointment reminders” to something more careful: “Some clinic visitors report difficulty completing required preparation steps before appointments, with the obstacles appearing to vary across document access, scheduling, and communication.” This is still an illustrative classroom statement, not a finding from actual clinic research. But notice the intellectual movement. Elena is now willing to describe uncertainty rather than hide it, because she understands that a precise question is more useful than an attractive but unsupported answer.

2.2.5.3 A problem statement can be both human and analytically rigorous

There is a common misunderstanding that academic accuracy requires emotionless language. In fact, a strong problem statement can communicate why something matters to a person's life while remaining careful about evidence. Consider a student who cannot arrange academic advising without missing paid work. The difficulty isn't merely an inconvenience in a calendar. It may affect income, confidence, and the ability to make timely academic decisions. Recognizing those consequences gives the inquiry a human purpose. But the investigator must still discover whether those consequences actually occur and which students experience them.

A well-written problem statement therefore combines two kinds of respect. It respects the person by recognizing the significance of their experience, and it respects the evidence by refusing to claim more than has been learned. These qualities support each other. Exaggerating a problem in the name of helping people may ultimately produce a poor solution that wastes resources or ignores the needs of the people it was supposed to serve. Describing it carefully creates room for people to correct the account, share overlooked details, and participate in evaluating possible improvements.

At the end of this section, we have moved beyond collecting stories into forming an initial explanation. The explanation remains provisional, because even an organized set of interviews is only one part of understanding a problem. The next task is to decide how much confidence the evidence deserves. What have we learned about frequency and intensity? Which voices may be absent? What else could explain the observations? And most important for an aspiring entrepreneur: even if the problem is real, have we learned anything yet about whether a particular solution would work?

Interactive Experiential Tool

Problem Statement Builder

For a workplace improvement or a startup idea. Your entries stay in this browser on this device; nothing is sent anywhere.

Build a working problem statement from your conversations or observations. The grey examples come from section 2.2.5; replace them with your own. A good statement names a group, a difficulty, a context, and a consequence, and leaves the solution open.

Your problem statement
Evidence note

Your working statement

Your statement appears here as you type.

Talk it through with the Module 2 AI Advisor

Emerging practices

Staying on Top of Emerging Practices

Organizations increasingly use AI systems to cluster interview notes, generate draft personas, and summarize thousands of customer comments. These tools can reduce clerical work and help investigators notice possible themes, but they can also convert weak assumptions into highly polished narratives. A generated persona may appear convincingly human without representing a single interview; a theme may sound frequent because a model repeats or merges similar language. Responsible analysis requires secure handling of data, careful review of original accounts, and explicit disclosure when a profile is hypothetical or AI-assisted. The central standard does not change with the technology: a compelling representation of a customer is not the same thing as evidence from a customer.

References

Sources cited in this section. The full Module 2 reference list closes section 2.3.

Blank, S. (2013). Why the lean start-up changes everything. Harvard Business Review, 91(5), 63–72. https://hbr.org/2013/05/why-the-lean-start-up-changes-everything

Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa

Brinkmann, S., & Kvale, S. (2015). InterViews: Learning the craft of qualitative research interviewing (3rd ed.). SAGE Publications.

Christensen, C. M., Hall, T., Dillon, K., & Duncan, D. S. (2016). Know your customers' “jobs to be done.” Harvard Business Review, 94(9), 54–62. https://hbr.org/2016/09/know-your-customers-jobs-to-be-done

IDEO. (2008). Analytics and inspiration. https://designthinking.ideo.com/blog/analytics-and-inspiration

Shepherd, D. A., & Gruber, M. (2021). The lean startup framework: Closing the academic–practitioner divide. Entrepreneurship Theory and Practice, 45(5), 967–998. https://doi.org/10.1177/1042258719899415