Skills Track: Every numbered section in this chapter belongs to the complete foundational entrepreneurship curriculum. The optional Venture Track companions, added after the academic manuscript was developed, show how an aspiring founder might apply the ideas to launching a business. You can skip those boxes without losing the academic lesson. Formal assignments and grading instructions belong in the separate Instructor Teaching Guide.
Learning Objectives
By the end of this module, you should be able to explain why recognizing a problem is not the same as understanding it and why entrepreneurial learning begins with the people who actually experience a difficulty. You will learn to ask questions that invite honest stories rather than polite agreement, listen for how people behave and what they have tried, and recognize when your own expectations may be shaping what you hear. You will also explore the ethical responsibilities involved in gathering information from others, including permission, privacy, accessibility, and respect.
You should be able to distinguish among users, customers, purchasers, and other people affected by a decision; identify meaningful customer segments; interpret recurring patterns and important exceptions in interview information; and describe a problem using evidence rather than guesswork. Finally, you will learn why understanding a genuine problem does not prove that a particular solution—or a profitable business—will work. These capabilities matter whether you plan to launch a venture, improve a workplace, enter a profession, or serve a community. Optional Venture Track companions will demonstrate practical startup applications of what you have learned.
Opening: The Problem Looked Obvious—Until Someone Asked
In Module 1, we met Elena, a fictional community-college student who works part-time in a community service setting. She noticed that people sometimes arrived at a clinic without the documents they needed for their appointments. Staff members spent time explaining instructions again, appointments ran behind, and people occasionally had to return another day. Elena recognized that a familiar frustration might represent an opportunity to improve a service. Instead of accepting the delays as normal, she asked why they happened. That was meaningful progress: she had begun to see problems through an entrepreneurial lens.
But opportunity recognition carries a subtle danger. Once we notice a problem, we often become attached to our first explanation. Elena believes the reminders are too confusing. The answer, she thinks, is to rewrite the reminders in simpler language. The idea sounds sensible, compassionate, and inexpensive. She can already imagine a better-looking message and the relief it might bring to busy people. There is just one difficulty: she has not yet asked the people who experience the problem what actually happens before their appointments.
Imagine that Elena starts a conversation by saying, “Our reminders are pretty confusing, aren't they?” A visitor might agree to be polite. Another might say that clearer reminders sound wonderful. Elena could return to the office with several encouraging comments and feel that her idea had been validated. Yet neither response tells her why a particular person arrived without documents. Perhaps the instructions were unclear; perhaps the documents were hard to obtain; perhaps transportation disrupted the trip; perhaps a patient understood the reminder but could not afford to take another morning off work. All these possibilities could produce the same visible outcome at the front desk.
Now imagine Elena approaches the conversation differently: “Could you walk me through what happened when you prepared for today's visit?” The question makes room for a story rather than pushing people toward her explanation. A visitor might describe needing to print a document but having no printer. Another might mention that the reminder arrived after the person had left for a long work shift. A third might explain that the instructions were easy enough, but an online form was inaccessible on an older phone. These are illustrative possibilities, not actual research findings. They demonstrate how quickly the meaning of a problem can change when the person affected has room to explain it.
Chapter 1 taught us to see opportunities. Chapter 2 teaches a harder and more generous skill: to make our own ideas less important, temporarily, so that other people's experiences can teach us something we do not yet know. This is the foundation of customer discovery. Although the term includes the word customer, the method is useful to students preparing for healthcare, education, skilled trades, social services, management, public administration, and many other professions. Wherever people create value for other people, understanding the experience of those affected is part of doing the work well.
Customer discovery begins with a recognition that should be familiar after Module 1: our observations matter, but they do not automatically explain themselves. The task now is to replace confident guessing with respectful inquiry. This is not always easy. We like solving problems, and an idea that feels generous or clever can become part of how we see ourselves. Discovering that the problem is different from what we imagined may feel disappointing. In this section, we will learn why that discomfort can be useful and how a conversation can become an instrument of learning rather than a performance of being right.
2.1.1 Why Customer Discovery Matters
2.1.1.1 The dangerous comfort of knowing what people need
Imagine spending weeks creating the perfect set of instructions for a workplace process. You select the clearest fonts, develop a logical sequence, include diagrams, and proudly distribute the guide. A month later, employees are still making the same mistakes. It is tempting to decide that they did not read the instructions carefully enough. But perhaps the guide is not the problem. The software may require an unusual sequence of steps, staff may receive conflicting directions from supervisors, or the task may change when the office gets busy. If you investigate only whether people like your guide, you may miss what makes the work difficult in the first place.
A familiar saying in entrepreneurship describes the mistake as building something and assuming people will come. The business version can be expensive: someone develops a product, rents a location, purchases inventory, or hires employees before understanding whether the intended customers experience the problem as expected. The organizational version can be just as frustrating: a manager introduces a new policy that adds work without addressing the underlying obstacle. In either setting, people have acted on an explanation that was never adequately examined. The risk is not simply that a solution fails; it is that effort and trust are spent trying to solve the wrong problem.
Steve Blank (2013), a central figure in the customer-development movement, argues that young ventures should treat their early ideas as assumptions to investigate rather than instructions to execute. Shepherd and Gruber (2021) connect customer development with a broader entrepreneurship research framework, explaining that learning about customers is part of searching for a workable business model. Neither source implies that every conversation can eliminate uncertainty or that interviewing guarantees success. The more modest, useful lesson is that informed questions can uncover important discrepancies before a person becomes deeply committed to a plan.
For a student who hopes to become an entrepreneur, discovery can prevent avoidable commitments. For a student who intends to become a nurse, mechanic, teacher, office manager, or engineer, it offers a different advantage: an approach to diagnosing work problems before prescribing interventions. Think about a technician who is repeatedly called back to repair what seems to be the same failure. A quick fix may restore operation temporarily. Understanding when the failure occurs, how the equipment is used, and what users have already tried may reveal the real cause. The professional's confidence grows not from guessing faster but from asking questions that expose missing information.
2.1.1.2 Discovery is not marketing, selling, or conducting a popularity contest
The word customer may make this activity sound like an early sales conversation. But discovery and selling are different processes. Selling attempts to persuade someone to choose an offering. Discovery attempts to understand the person's situation well enough to learn whether an offering is even appropriate. Both may eventually matter to a venture, but mixing them too early creates problems. A person who senses that you want approval may offer encouragement instead of describing a difficult experience honestly. A person who believes you are about to sell something may become defensive or withhold useful details.
Customer discovery also differs from a simple satisfaction survey. A survey might ask 200 people to rate a service on a scale from one to five. That can be useful for measuring a defined experience when the questions and sampling approach are appropriate. A discovery interview asks a different kind of question: What was happening when the service became difficult, and how did you handle it? The answer may reveal a sequence of decisions, an unexpected constraint, or a workaround the survey designers never thought to include. One form of information is not universally superior. The methods answer different questions, and responsible investigators select the method that fits what they need to learn.
The distinction becomes especially important when people say they like an idea. Imagine a college employee proposing a new mobile app for academic advising. Students may respond warmly because the concept sounds convenient. But the institution still does not know whether students already use another tool, whether they have reliable internet access, whether advisers can maintain the information, or whether an app would solve the most consequential difficulty. A favorable opinion is an observation about a conversation; it is not evidence of actual future use. Research on lean startup practices stresses the value of hypothesis-driven probing while also showing that implementation and results depend on context (Leatherbee & Katila, 2020).
A useful way to recognize the difference is to ask yourself what you want the person to say before the conversation begins. If the hoped-for answer is “That's a wonderful idea,” you may be preparing to pitch rather than investigate. If the hoped-for outcome is a more accurate understanding—even when the answer challenges your idea—you are moving toward discovery. This does not mean abandoning enthusiasm or treating people like sources of data to be mined. It means respecting their experiences enough to listen without forcing them to confirm what you already believe.
2.1.1.3 Discovery as a habit of professional service
We can understand discovery as a form of professional humility. A good instructor does not assume every student who misses an assignment is careless. A responsible healthcare worker does not assume every missed appointment reflects a lack of concern. A skilled tradesperson does not assume a customer's account of a problem is technically complete, yet does not dismiss the customer as irrelevant either. Each professional combines expertise with inquiry, recognizing that technical knowledge and lived experience illuminate different parts of the situation. The people who experience a process know things its designers may never see.
This matters in institutions that have been operating for many years. Employees may become experts at the existing process and, in doing so, lose sight of what it feels like to encounter the process for the first time. Imagine navigating college registration without knowing which words in a catalog have technical meanings. To a registrar, prerequisite, corequisite, and program map may be routine. To a first-generation student, those terms may be unfamiliar, and the difference could affect an entire semester of decisions. Discovery helps organizations regain the perspective of people whose experiences are not already shaped by inside knowledge.
A responsible investigator must also understand limits. There are situations in which extensive prior evidence, professional standards, or safety requirements already dictate the correct action; asking customers whether they prefer a less safe procedure is not a substitute for following those requirements. In other circumstances, especially where experiences differ widely, respectful discovery can reveal information that experts need to design accessible services. The practical question is not “Should I listen instead of using expertise?” but “What do I know through expertise, what can affected people teach me, and how do I combine those forms of knowledge responsibly?”
2.1.2 Empathy and the Customer's Perspective
2.1.2.1 Empathy is curiosity with boundaries
Empathy is sometimes portrayed as the ability to feel exactly what another person feels. That description sounds admirable, but it can make us overconfident. We cannot fully inhabit a life we have not lived, and even people with similar backgrounds may experience the same event differently. In customer discovery, a more useful understanding of empathy is the deliberate effort to learn how another person experiences a situation, while recognizing the limits of our understanding. Empathy begins by taking the person's account seriously, not by announcing that we already know how they feel.
Consider a worker who must attend a required training session at 8 a.m. A manager might interpret absence as resistance to professional development. An employee might experience the schedule as a conflict with childcare or public transportation. Another might be available at 8 a.m. but unable to complete the online preparation because a home internet connection is unreliable. An empathetic investigator does not assume every scheduling difficulty has the same cause. They ask about circumstances, consequences, and constraints with enough respect that people can explain their experience without being judged for it.
Researchers and practitioners increasingly connect empathy with innovation, but the concept should not be used as a flattering label for intuition. Christensen, Hall, Dillon, and Duncan (2016) argue that understanding the progress people are trying to make can be more useful than describing customers only through demographic categories. Their Jobs to Be Done perspective asks what a person is trying to accomplish in a particular situation. An employee is not necessarily asking for more training options; they may be trying to meet a professional requirement without sacrificing essential caregiving responsibilities. Understanding that larger goal can open possibilities a narrow focus on the existing training schedule would miss.
Empathy also imposes a responsibility not to turn someone's difficult experience into an inspiring story without permission. People may discuss disability, debt, immigration status, family stress, or other sensitive matters when explaining a practical challenge. Their willingness to speak does not make those details available for public sharing. Good discovery pays attention to the person as a whole human being with choices and boundaries. The fact that their difficulty might inspire an improvement or a business idea does not entitle an investigator to claim ownership of their story.
2.1.2.2 Context changes the meaning of behavior
Suppose two students never attend in-person tutoring sessions. A quick conclusion might be that both prefer studying independently. But one student may work during the tutoring center's hours, while the other may feel uncomfortable asking for help in a busy room. Their visible behavior is the same; its meaning differs. This is one reason context matters. Context includes the conditions in which a person acts: time, place, access to resources, relationships, rules, previous experiences, culture, and obligations. Without context, we may read a person's choices as preferences when those choices are actually responses to constraints.
The distinction is especially important when observing money and time. Someone who purchases an expensive convenience product may not be careless with money; they may face a schedule that makes a slower alternative impossible. Someone who uses a complicated workaround may not enjoy doing so; the workaround may simply be the only available option. Conversely, an inconvenient process may be acceptable to someone who values reliability or privacy more than speed. An investigator should not treat every friction point as an opportunity to optimize a single measure, such as minutes saved or clicks reduced. Meaningful value is defined partly by what people are trying to protect and achieve.
In the fictional clinic scenario, Elena's original idea was to improve reminder language. As she explores context, she realizes that a reminder is just one event in a larger chain: the person receives a message, interprets it, locates documents, rearranges work or care responsibilities, travels, and checks in. A process may fail at any step. If she studies only the message, she may miss the point where people actually lose the ability to complete the task. Mapping the experience doesn't require assuming everyone follows one standard path; it requires asking people to describe their own sequence and identifying where meaningful differences occur.
This ability carries into nearly any profession. A supervisor evaluating a new shift system must understand workers' circumstances. A librarian considering longer hours must learn when patrons can reach the library and what services they need. A mechanic explaining a recurring repair should ask how and when the vehicle is used, even while relying on professional diagnostic procedures. In each case, empathy broadens the investigation from “What do I think people should do?” to “What conditions shape what they can actually do?”
2.1.2.3 Jobs to Be Done: what progress is the person seeking?
The phrase Jobs to Be Done can sound as though it refers only to employment. In this framework, a job means the progress a person wants to make in a situation. Someone buying a suitcase may want to protect belongings and travel without stress. Someone seeking a tutoring session may want to regain confidence before an examination. Someone arranging home maintenance may want safety, predictability, and the freedom to focus on other obligations. The product or service is a possible means of achieving the goal, not the goal itself.
Consider the difference between asking, “Why aren't students using our appointment app?” and “How do students make sure they get the advising help they need?” The first question takes an existing solution as the starting point. The second opens the investigation to the larger task. Students may rely on classmates, send emails, visit offices between jobs, use printed handouts, or postpone a decision until a deadline. Each behavior may reveal something about what a workable service would need to accomplish. Some students may consider a mobile app helpful; others may prefer a knowledgeable person because their questions are complicated. Discovery asks about the progress, not merely the channel.
Christensen et al. (2016) use the Jobs to Be Done perspective to challenge organizations that confuse detailed customer data with a clear understanding of why customers choose an offering. That argument doesn't imply that age, language, income, disability, or other contextual characteristics are irrelevant. Such factors can shape access, needs, and constraints. The point is that demographic descriptions alone rarely explain a particular decision. To understand the decision, we need to know what mattered in the situation and which alternatives the person considered.
Elena can apply this by asking what visitors were trying to accomplish when preparing for the clinic. The answer may seem obvious—attend an appointment—but the meaningful progress could include obtaining care without missing paid work, avoiding embarrassment, protecting privacy, or bringing a family member for support. Different priorities can produce different responses to the same reminder. By listening for goals and constraints, Elena begins to see that a service is experienced as an entire journey, not as one message created by the organization.
2.1.3 Asking Better Questions
2.1.3.1 An interview is a conversation with a learning purpose
Most of us already know how to have a conversation. We ask what happened, offer sympathy, tell related stories, and sometimes recommend a solution. A customer discovery interview uses some of those familiar habits but changes the purpose. The interviewer is not primarily trying to be entertaining, win approval, or deliver an impressive idea. The interviewer is trying to learn what actually happens in a person's life or work. That requires preparation, careful listening, and an unusual willingness to leave silence unfilled.
An interview can be informal in tone while still being systematic in design. Before the conversation, the investigator should identify what they genuinely need to learn: how often a difficulty occurs, what the person did most recently, which alternatives were available, and what consequences followed. An interview guide is a short list of open questions and follow-up prompts, not a script that must be read word for word regardless of what the participant says. A good guide creates enough consistency to compare experiences while leaving room for unexpected details. Brinkmann and Kvale (2015) emphasize interviewing as a craft that involves decisions about meaning, context, ethics, and the quality of what is learned.
The difference between an effective and ineffective question is often subtle. “Don't you think this process wastes too much time?” contains the preferred conclusion. “How long did the process take the last time you used it?” invites a concrete account. “Would you use a convenient app?” makes it easy to say yes without considering an actual choice. “What did you use when you needed help last week?” points toward behavior. Questions about actual events are often more useful than predictions about hypothetical futures, although they still require interpretation and cannot reveal every circumstance.
The entrepreneur and author Rob Fitzpatrick popularized this principle in The Mom Test (2013): even a supportive parent can give you a more useful answer when you ask about their own experiences rather than inviting them to judge your cherished idea. The lesson is not that people are dishonest. Most people are trying to be kind, cooperative, or responsive to the conversation in front of them. Our responsibility as interviewers is to ask questions that do not make politeness the easiest route through the discussion.
2.1.3.2 Past behavior, concrete events, and gentle follow-up questions
A useful opening question often begins with a recent event: “Tell me about the last time you tried to schedule an appointment.” This is stronger than “Do you have problems scheduling appointments?” because it invites a sequence. What made the person decide to schedule? Where did they look? What happened first? What did they try when a step was difficult? How did the situation end? Each answer can lead to a neutral follow-up: “What happened next?” or “Could you explain that part?” The conversation begins to resemble a story with actions and constraints rather than a collection of general opinions.
There are limits to this approach. Memory is imperfect, and a person may remember an unusual recent experience more vividly than routine ones. Some events are too sensitive to discuss casually. A participant may decline to answer, and the investigator should not pressure them. When a person describes a behavior, another useful question is “Would you be comfortable showing me how that process works?” Demonstration can reveal steps that are difficult to remember or explain. But permission and context matter: no one should be asked to reveal private accounts, medical information, passwords, or proprietary work materials to satisfy a student's curiosity.
Consider this fictional exchange. Elena asks, “Tell me about preparing for your last appointment here.” A visitor replies, “I kept putting it off until the night before.” Rather than announcing, “So you need more reminders,” Elena could ask, “What made it difficult to prepare earlier?” The visitor might mention a rotating work schedule. Elena could continue, “What did you try when you found out you needed a printed form?” The visitor might explain that a neighbor eventually helped. The workaround—seeking help from another person—becomes a clue about access and timing, not proof that rewriting the message is the correct intervention.
A comparison helps reveal the principle:
| Question that pushes toward an answer | Question that makes room for learning |
|---|---|
| “Isn't our scheduling system confusing?” | “Walk me through your most recent scheduling experience.” |
| “Would you love a simpler reminder?” | “What did you do after you received the reminder?” |
| “How much would you pay for an app?” | “What, if anything, did you spend to handle the problem last time?” |
| “Wouldn't evening hours solve this?” | “What makes available appointment times workable or difficult for you?” |
| “Was the employee unhelpful?” | “What help did you ask for, and what happened next?” |
These alternatives are not magic phrases. A neutral-sounding question can still be delivered in a pressuring tone, and a person can still try to guess what answer is expected. Good interviewing requires attention to posture, pacing, context, and the relationship between the people involved. The strongest signal of curiosity may be how the interviewer responds to a surprising answer: do they explore it, or do they steer the conversation back to the idea they wanted to confirm?
2.1.3.3 Listening without taking over the story
Imagine explaining a frustrating experience to someone who interrupts after every sentence with suggestions. The suggestions may be well intended, but eventually you may stop describing what happened and simply respond to the other person's ideas. The conversation has changed from learning to advising. A discovery interviewer faces the same temptation, especially when a solution seems obvious. The ability to listen is not passive. It involves choosing what not to say so another person can finish an account that may challenge your assumptions.
Active listening includes attentive body language where culturally appropriate, brief acknowledgments, requests for clarification, and checks on understanding. A researcher might say, “So the appointment time itself worked, but the paperwork was difficult to obtain—is that accurate?” The person can confirm, reject, or revise the interpretation. This is more respectful than silently recording a conclusion as though the interviewer knows what the experience meant. It also helps distinguish a participant's words from the analyst's explanation, a distinction we will return to when analyzing evidence.
Sometimes the most revealing information arrives after a pause. A person may add, “Actually, the hardest part wasn't the form,” or “I didn't mention that I share a phone with a family member.” A hurried interviewer may miss these details because they are focused on reaching the next scripted question. A careful interviewer knows the topic they need to explore but remains open to a conversation that develops differently than expected. That balance—structure without rigidity—is one of the skills that turns casual discussion into a useful inquiry.
The transferable lesson is larger than customer research. An effective supervisor listens before changing a process. A teacher listens before deciding a student is unmotivated. A social-service professional listens before concluding that a client has simply failed to follow instructions. Discovery can improve the quality of decisions because it turns people from objects of our analysis into partners in understanding a situation. Their stories still must be interpreted and checked, but they should not be replaced by the investigator's imagination.
Entrepreneur-in-Action
Intuit: Learning by Seeing the Work People Actually Do
Intuit, the company behind financial software products such as TurboTax and QuickBooks, has publicly described a customer-research practice called Follow-Me-Home. The company asks teams to observe customers using tools or performing tasks in real contexts, with their permission, rather than relying exclusively on what customers remember or what designers imagine. Intuit's own published method materials emphasize that a follow-me-home is primarily an observation exercise, not a product pitch or a conventional interview (Intuit, n.d.). This is important because a customer who says a process is simple may still reveal, through a demonstration, several unexpected steps and workarounds.
The company has also described employees engaging directly with people preparing their taxes. In a 2017 account, Intuit explained that learning about people's broader lives and circumstances helped inform decisions about how TurboTax was presented and personalized (Intuit, 2017). Those reports are company accounts of its own practices, not independent proof that observation alone produced a particular commercial outcome. Yet the documented behavior illustrates a principle that students can use in many environments: a service becomes easier to understand when we see the conditions under which people actually use it.
Imagine a student interested in improving a campus computer lab. Asking, “Do you find the lab helpful?” may yield little detail. With permission and without collecting private information, observing how newcomers locate equipment, sign in, and ask for help may reveal a more useful picture. A person may hesitate at a sign, repeat a step, or ask another student for assistance. Those moments are not automatic instructions to redesign the facility, but they give the investigator better questions to ask. The lesson from Intuit is not to imitate an expensive corporate research program. It is to recognize that people's real behavior can teach us things polished explanations often leave out.
Recruit Five Relevant Potential Customers
Imagine Sam wants to start a mobile bicycle-repair service for people who use bicycles to commute to work. He could ask five friends whether mobile repair sounds convenient. But if those friends rarely ride a bicycle to work, he would learn more about politeness than about the problem. Sam's first tactical decision is whom to approach: people who have recently experienced a breakdown, maintenance delay, or difficulty getting a bicycle to a repair shop. Five conversations are a practical first learning step, not a universal validation threshold.
A founder's respectful recruitment message: “I'm exploring how people who commute by bicycle handle unexpected repair and maintenance problems. I'm not selling anything. Would you be open to a 10–15 minute conversation about a recent experience? Participation is completely optional; I won't record or share identifying details without your permission.” Sam would use appropriate, permitted channels, respect refusals, and avoid implying that the participant has any obligation to help.
A simple five-question discovery guide: (1) “Tell me about the last time your bicycle needed repair.” (2) “What were you trying to get done that day?” (3) “How did you handle the problem?” (4) “What did that require in time, effort, or money?” (5) “What, if anything, would you do differently next time?” Useful neutral follow-ups include “What happened next?” and “Could you give me an example?” He would avoid mentioning a mobile service until the experience was understood.
Founder caution
Do not approach people in situations where they may feel pressured to participate. Do not solicit minors, sensitive records, employer-confidential information, or recordings without appropriate safeguards and permission. Five relevant interviews can expose bad assumptions, but they do not establish market size or customer demand.
What this teaches
Recruit for relevant experience, explain the conversation honestly, and use a neutral guide before pitching a product.
2.1.4 Research Ethics and Participant Respect
2.1.4.1 Asking a good question does not give you the right to an answer
We have spent time learning how to ask questions that reveal experience. Now we must consider whether, when, and how those questions should be asked at all. A person's willingness to be helpful is not unlimited permission to collect information about their life. Research ethics begins by recognizing that the person is not simply a source of useful details. They have an independent right to decide whether to participate, what to share, and whether the conversation should continue. This is especially important when the person may feel pressured by an employer, teacher, healthcare provider, or someone who controls access to a service.
Informed participation means explaining what you are trying to learn, what participation involves, how you intend to use the information, and that declining is acceptable. You should not suggest that a conversation is an official survey, clinical assessment, or approved research study when it is not. An ethical introduction could be as simple as: “I'm learning about difficulties people experience with scheduling. Would you be willing to discuss your own experience for a few minutes? You can skip any question or stop whenever you want.” In a real workplace, an employee may still need managerial approval or formal research review before contacting customers. Good intentions do not bypass the policies that protect them.
The ethical principles described in the U.S. Belmont Report—respect for persons, beneficence, and justice—are foundational in human-subject research (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979). A casual classroom interview is not automatically a regulated human-subject study, and it is important not to confuse those settings. Nevertheless, the underlying considerations are useful: respect choice, consider possible harms, and avoid placing the burdens of participation unfairly on particular groups. When an activity involves a vulnerable population, sensitive data, or an institutional setting, the instructor or organization should determine what permissions or safeguards are required before anyone begins.
Elena's clinic provides a clear example. She should not independently interview patients about medical situations, obtain records, or make promises about the clinic's privacy protections. She would need appropriate authorization and may be better served by using a fictional classroom case, an approved de-identified process study, or a professional's general explanation of existing procedures. The entrepreneurial principle still applies, but a desire to improve a service must never outrank patient confidentiality or ethical obligations. Some of the most professional decisions involve recognizing that a particular research method is inappropriate and choosing a safer one.
2.1.4.2 Privacy, confidentiality, and data minimization
Privacy concerns are often discussed as though they belong only to large technology companies. Yet a student with a notebook or a phone can collect information that a person would not want shared. An interview about commuting may reveal a home address, employment schedule, or family obligation. A conversation about finances may reveal debt, income insecurity, or other sensitive circumstances. The fact that information is easy to record does not mean that keeping it is useful or justified.
A practical principle is data minimization: collect only what you need for the clearly stated learning purpose. If your research question concerns how students experience advising appointments, you may need to know whether a person works during business hours. You probably do not need their employer's name, exact income, or full class schedule. If you can learn from an anonymized summary, do not retain identifying details simply because the participant happened to mention them. Secure storage, limited access, and a plan for deleting unnecessary notes are as important for a small project as they are for a larger one.
It is also worth distinguishing anonymity from confidentiality. An anonymous record cannot readily be connected to an identified individual, while confidential information may be identifiable but is protected from unauthorized disclosure. If you personally know whom you interviewed, you should be careful about claiming the interview was anonymous. Recording audio creates a particularly sensitive record because a voice may identify a person and the conversation may contain details not needed for analysis. Recording should require explicit permission; a person's agreement to talk does not automatically include agreement to be recorded or uploaded to an AI service.
Good ethics can improve research quality as well as protect participants. People who understand why they are being asked and believe their choices will be respected may be more willing to share candid experiences. But ethics is not merely a technique for getting better answers. Respect is an obligation even when it reduces the amount of information we can collect. If someone chooses not to discuss a difficult experience, the appropriate response is to accept the boundary—not to label the person an unhelpful customer.
2.1.4.3 Accessibility, power, and who gets heard
Imagine an organization asking for feedback through an online form available only in English and requiring a current smartphone. The organization may receive many responses and conclude that it understands its community. But who was able to participate? People with limited digital access, different language needs, or certain disabilities may be missing. The absence of those voices can make a service appear more inclusive than it really is. A responsible discovery process considers not only what people say but also how the method selects which people get the chance to speak.
Accessibility involves more than providing a convenient interview time. Some participants may need an interpreter, a quiet environment, a written explanation, or an alternative way to contribute. Others may prefer not to disclose personal difficulties to someone connected with an institution that serves them. Social power matters: a worker may hesitate to criticize an employer's policy when a supervisor is present, and a student may feel pressure to praise a professor's program. These relationships can shape both participation and answers. A thoughtful interviewer seeks to reduce pressure and recognizes that a polite response may reflect the circumstances of the conversation.
Good discovery also avoids treating people who are less easily reached as less important. Sometimes those facing the greatest difficulties are also the least available for research. A student working two jobs may not attend an afternoon focus group; a rural resident may not have dependable internet; a caregiver may not be able to schedule a lengthy conversation. Research design should not automatically equate convenience with relevance. When some perspectives cannot be included, that limitation should be acknowledged rather than hidden beneath a confident claim to speak for everyone.
2.1.5 Recognizing Bias in Discovery
2.1.5.1 The answers we hope to hear can shape the questions we ask
Most interview mistakes do not begin with deliberate dishonesty. They begin with a familiar human tendency: we notice information that fits what we already believe. This tendency, called confirmation bias, can affect whom we recruit, what questions we ask, which parts of a conversation we remember, and how we summarize findings. Elena might ask five visitors whether reminders should be clearer, hear agreement, and return with the conclusion that reminder language is the clinic's major problem. She may have collected responses accurately but asked a question too narrow to test her underlying explanation.
Confirmation bias is not solved simply by writing “be objective” at the top of an interview guide. The investigator needs a method for inviting information that could challenge the preferred account. Elena might ask about the last appointment preparation experience and follow that with “What was easier than expected?” or “Were there any other reasons the visit was difficult?” She might deliberately seek stories from people who arrived prepared as well as those who did not. If one visitor describes a problem unrelated to reminders, that account deserves attention rather than being discarded because it complicates her idea.
Courtesy bias occurs when people give answers that seem polite or agreeable rather than expressing their full opinion. The interviewer's role, authority, or enthusiasm can increase this pressure. A participant may say “That's a great idea” when what they mean is “I appreciate your concern, and I don't want to disappoint you.” This is why a compliment should not be counted as equivalent to evidence of behavior. Fitzpatrick (2013) encourages founders to investigate people's actual experiences rather than solicit flattering judgments about their ideas. The same principle helps a manager evaluate a training proposal or a student compare professional-service options.
Some bias is created by what the researcher chooses to measure. If a service asks only how quickly a transaction was completed, it may miss whether the person understood the information or felt safe completing it. If an interviewer asks only about frustrations, they may overlook the features people value and would not want changed. Good discovery asks both what is difficult and what currently works. Understanding why an existing workaround is preferred can be as important as identifying why it is inconvenient.
2.1.5.2 A convenient sample is not the same as a representative community
Suppose you ask five friends whether they have difficulty finding affordable transportation to campus. Three describe serious problems. That is useful information about those three experiences, but it does not establish that 60 percent of students have the same problem. Your friends may share your class schedule, neighborhood, age, or access to a car. They may be more willing to speak with you than strangers would be. The group is a convenience sample, selected partly because its members were easy to reach. Its experiences deserve respect, but they cannot automatically be generalized to a much larger population.
Another approach is purposive sampling, in which researchers intentionally seek people with experiences relevant to a defined question. If the question concerns evening students' access to services, it makes sense to include students who attend in the evening and perhaps staff who serve them. This does not turn a small purposive sample into a statistically representative survey. It makes the conversations more relevant to the topic being explored. Qualitative interviews are often valuable for understanding how a problem works and why experiences vary, not for calculating the precise percentage of an entire population that agrees with a statement.
Guest, Bunce, and Johnson (2006) studied thematic saturation in a specific qualitative research context. Their work is frequently cited in discussions of interview sample sizes, but it should not be reduced to the claim that a fixed number of interviews is always enough. The amount and diversity of information needed depend on the people involved, the complexity of the question, the purpose of the research, and the degree of variation in experiences. Five thoughtful interviews can expose flaws in an initial assumption. They do not automatically validate a market or reveal everything an investigator needs to know.
For an entry-level student, the practical skill is learning to say what a set of conversations can and cannot support. “Three of the five people I interviewed described trouble accessing evening advising” is a transparent summary. “Most students cannot access advising” is a much broader claim requiring stronger evidence. The first statement can motivate further investigation. The second, if unqualified, may encourage decisions that do not fit the wider community.
2.1.5.3 When the surprising answer is the most useful answer
Imagine Elena has spent an afternoon listening to people who arrived without documents. She expects to hear that reminders are unclear. Instead, two people report that the reminders were understandable but the required forms were available only during business hours. Another says the instructions were clear but the transportation connection caused them to arrive too late for staff assistance. A fourth says the reminder never arrived. These fictional answers do not form a statistically reliable study. But they show that “confusing reminder” may be too narrow a description of the broader preparation problem.
At this point, Elena faces a choice. She can dismiss the unexpected accounts as exceptions and keep pursuing the idea she likes. Or she can ask whether the differences reveal several mechanisms that produce the same visible problem. The second response is intellectually harder, because it delays the satisfaction of being able to say, “I found the answer.” It is also more consistent with the mindset developed in Module 1: curiosity, reflection, and the willingness to revise an assumption when the evidence demands it.
The stronger investigator therefore develops a habit of documenting disconfirming evidence—information that challenges the leading explanation. One thoughtful exception may not overturn a larger pattern, but it can reveal that a conclusion needs to be narrowed. A person who arrived prepared despite facing the same constraints may also teach something important. What made their experience different? What supports or alternatives did they have? An inquiry that includes exceptions often produces a better understanding than one that searches only for a recurring complaint.
The aim of Module 2 is not to train you to distrust everything people say. It is to help you treat their accounts with seriousness while recognizing that memories are incomplete, circumstances vary, and your questions influence what you learn. Once you can hold all three ideas together, customer discovery becomes more than collecting encouraging anecdotes. It becomes a disciplined way of improving judgment under uncertainty.
Emerging practices
Staying on Top of Emerging Practices
Digital services can now provide click paths, call transcripts, search logs, and AI-generated summaries that appear to reveal customer behavior at extraordinary scale. These tools can identify where people leave a process or which questions are repeated, but they do not automatically explain why those things happen. A click may represent confusion, a deliberate choice, a technical error, or an accessibility barrier; an AI-generated theme may reflect a biased or incomplete dataset. As customer research becomes increasingly automated, the essential skill remains asking what a record actually represents, which people are absent from it, and whether the organization has permission to collect and use it. Responsible tools should expand human understanding rather than replace human voices with convenient assumptions.
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
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
Fitzpatrick, R. (2013). The mom test: How to talk to customers & learn if your business is a good idea when everyone is lying to you. CreateSpace Independent Publishing Platform.
Guest, G., Bunce, A., & Johnson, L. (2006). How many interviews are enough? An experiment with data saturation and variability. Field Methods, 18(1), 59–82. https://doi.org/10.1177/1525822X05279903
Intuit. (n.d.). Design for Delight methods: Deep customer empathy [Method cards]. https://www.intuit.com/content/dam/intuit/intuitcom/partners/documents/education/icom-edu-d4d-method-cards.pdf
Intuit. (2017, February 17). Intuit named to Fast Company's top 10 companies in design list for 2017. https://investors.intuit.com/news-events/press-releases/detail/387/intuit-named-to-fast-companys-top-10-companies-in-design-list-for-2017
Leatherbee, M., & Katila, R. (2020). The lean startup method: Early-stage teams and hypothesis-based probing of business ideas. Strategic Entrepreneurship Journal, 14(4), 570–593. https://doi.org/10.1002/sej.1373
National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. (1979). The Belmont report: Ethical principles and guidelines for the protection of human subjects of research. U.S. Department of Health and Human Services. https://www.hhs.gov/ohrp/regulations-and-policy/belmont-report/read-the-belmont-report/index.html
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