Key Takeaways
Treat customer conversations as a way to find work worth building
Listening prevents waste. Michele Hansen began interviewing customers after seeing teams rely on analytics and educated guesses to decide what to build. Numbers can reveal what customers do, but not why. A conversation can expose a workaround, a hidden goal, or a reason a promising feature would go unused. Hansen credits customer listening as a core operating practice at Geocodio, the self-funded software company she co-founded, which surpassed $1 million in annual revenue. That is an example of the practice embedded in a business, not proof that interviews alone caused its success.
Make it routine. Interviews can inform product decisions, pricing, marketing, and retention. Their value is not limited to startups: a small company can use them to avoid months of wasted development, while a larger organization can bring engineers and other decision-makers closer to the people using its products.
What's valuable here is the shift from treating research as a special project to treating it as part of how a company thinks. Direct exposure to customer reasoning can correct the abstractions that accumulate in dashboards, roadmaps, and internal debate. The Geocodio example makes the practice concrete without establishing a clean causal link between interviews and revenue. That distinction matters: conversations are a way to improve decisions, not a guarantee of commercial success. Their strongest justification is often modest and practical: learn enough to avoid confidently building the wrong thing.
Trace the customer's whole process, not just your product's step
Find the larger job. A process is a set of steps taken toward a broader goal. Buying shampoo, for instance, may be part of someone trying to manage fine hair and feel good about their appearance. In a practice interview, Hansen's partner uncovered that her shampoo purchase was connected to a recent beach trip and a hope of recreating how her hair looked there. The shelf choice made more sense once the interviewer explored the goal and memory behind it.
Look beyond function. The same process has functional dimensions, such as getting a task done; emotional ones, such as feeling confident; and social ones, such as meeting others' expectations. A product usually occupies only one step. Laundry detergent pods illustrate how a small improvement, removing the measuring step, can be valuable even without replacing the entire process.
A process view helps teams escape feature-centered thinking. Customers experience outcomes, transitions, and workarounds, not a product's internal feature map. It also explains why superficially similar users may need different solutions: time pressure, social expectations, and emotional stakes alter what counts as useful. The shampoo story is especially effective because an apparently impulsive purchase turns out to have a personal logic. One caution is that a vivid explanation from one person is a hypothesis about meaning, not evidence that the same motivation is common across a market.
Slow your replies until customers reveal the reasoning beneath their actions
Make room to explain. The interviewer is not there to impress the participant with expertise or solve the problem on the spot. A gentle tone, attentive silence, and short acknowledgments make it easier for someone to keep speaking. Hansen recommends letting the customer do most of the talking, then mirroring a phrase or briefly summarizing it to invite more detail. A statement such as a neutral restatement can prompt elaboration without steering the person toward a preferred answer.
Listen without judging. Avoid interrupting, correcting, defending the product, or filling every pause with possible answers. Treat the customer as the authority on their experience, even when their facts or terminology are imperfect. The interviewer can assess accuracy and business implications afterward. In the moment, the task is to understand how the person's choices made sense to them.
The techniques work by changing the social incentives of a conversation. When an interviewer corrects or defends, the participant may start managing impressions rather than describing experience. Silence and reflective summaries communicate attention while leaving the participant in control of what to disclose. This is not passive listening: it requires active restraint and careful follow-up. The approach also calls for ethical limits. Making people feel unusually safe can lead them to reveal more than they intended, so respectful interviewing includes noticing discomfort and allowing a topic to change.
Ask what people did, not what they predict they might do
Prefer evidence over promises. Questions such as whether someone would use or pay for a proposed product invite guesses and politeness. Ask instead what they currently use, what they tried before, how long a task takes, how often it occurs, and what it costs in time or money. These concrete details can reveal willingness to pay more reliably than a hypothetical price question.
Unpack the workaround. A person may not call a process painful, yet still describe several manual steps, multiple tools, or days of effort. In the book's fundraising example, an employee spent up to a week combining data to choose event locations, for a process repeated annually and tied to substantial fundraising revenue. The current software cost alone would have understated the problem's value. Ask about consequences if the task goes undone, as well as the effort involved in completing it.
Past behavior reduces the burden of speculation and gives the interviewer observable anchors: a recent task, an existing expense, or a workaround someone has already chosen. That is a stronger basis for product decisions than enthusiastic agreement with an idea. Still, behavior is not a perfect forecast. A costly workaround may be tolerated for years without creating demand for a new product, and a customer may lack authority to buy. The interview therefore needs to uncover both the work itself and the context around adoption, budgets, and decision-makers.
Pair interviews with data, then narrow the question in five-person rounds
Use each method for its strength. Analytics and surveys can show how common a behavior is; interviews help explain what is behind it. Neither can replace the other. For a focused question, Hansen suggests starting with about five interviews, then reviewing what repeats and what remains unclear. Research studies she cites found that five interviews can reveal many usability needs, but more complex questions may require ten to fifteen people. Five is a practical starting point, not a universal proof threshold.
Refine in cycles. If five participants describe unrelated processes, the research scope may be too broad. Analyze the first set, choose a narrower audience or problem, and interview another group. Keep broader, ongoing conversations alongside targeted research on specific issues. Recruit a diverse range of participants, since excluding people with disabilities or other underrepresented groups can hide needs and business opportunities.
This approach balances momentum with humility. Small interview rounds are feasible for founders, while repeated narrowing prevents early exploration from turning into an unfocused collection of anecdotes. The key discipline is to distinguish learning a pattern from estimating its prevalence. Five conversations can surface mechanisms and questions; they cannot establish how many customers share a need. Combining interviews with behavioral data, surveys, or other evidence helps keep compelling stories from being mistaken for market-wide facts.
Recruit around real behavior, and keep research separate from selling
Find people close to the problem. Potential participants may already be discussing a frustration in a forum or social network; current customers can be selected by how recently they joined, what they pay, or which features they use. Ask about a specific, recent experience rather than inviting anyone with a vague interest. A clear request should explain who is seeking input, what the conversation involves, and how long it will take. An incentive may be appropriate when participation feels like work or recruitment is cold.
Protect the purpose. Customer research is distinct from support and sales. In support, first help the person resolve the issue. In a sales conversation, respect that the person is evaluating a product. In an interview, do not pitch, correct, or use a customer's answer to close a sale. If a relevant product feature comes up, follow up separately after the interview.
Recruitment is part of research quality, not merely logistics. Asking people who have just experienced a process improves the specificity of their recall, while careful targeting helps avoid conversations with people who cannot speak to the question. Separating research from sales also reduces pressure to perform enthusiasm or defend a purchase. Yet recruitment channels shape who volunteers: people eager to complain or unusually available may be overrepresented. Clear screening, varied outreach, and cautious interpretation help limit that distortion.
Test the buying journey, not just whether the product works
Observe the next step. An interactive interview puts a prototype, website, or existing product in front of someone and asks them to narrate what they expect and do. Resist the urge to guide them through a difficulty; that difficulty may be affecting many people who never contact support. Also ask what they would do after deciding the product seems useful, including whom they would consult and what approval is required.
Uncover hidden gates. Hansen recounts testing a healthcare data service that appeared valuable and usable. Participants then explained that using it required legal and security reviews that could take six months or longer. That made a low-cost, self-service plan commercially unworkable, despite the product's technical appeal. The company shifted to an enterprise model, and the product eventually found traction. The obstacle was not the interface but the customer's purchasing process.
This story exposes a frequent category error: confusing product usability with adoption feasibility. In organizations, decisions travel through procurement, security, legal review, and internal advocacy. A prototype test that stops at task completion misses those institutional mechanics. The lesson extends beyond software: ask what happens after a user says yes. The account also illustrates why an interview can reveal constraints but cannot settle the business model by itself. Commercial viability still depends on costs, sales capacity, and evidence from actual transactions.
Study satisfied customers before letting cancellations dictate the roadmap
Balance the evidence. Canceled customers can explain disappointment, but focusing only on them risks changing a product for people it was never meant to serve. Happy, long-term customers reveal which use cases the product already handles well and where similar customers might be found. Hansen describes this imbalance as loss aversion: the tendency to give losses more attention than equivalent gains.
Choose the customer fit. A pizza shop could respond to dissatisfied salad-seekers by expanding its menu, or focus on parents ordering pizza for a youth sports team, a group that returns and is already well served. The choice depends on strategy and resources, not on whose complaint is loudest. For software, compare the needs of customers who stay with the needs of those who leave, then decide which audience the company can serve effectively.
The reminder to interview happy customers counters a real organizational bias: visible dissatisfaction can dominate planning even when it comes from a poor-fit segment. Studying successful use cases supports retention and clearer positioning, rather than treating every departure as a product defect. The pizza example also shows that customer research does not remove strategic judgment. A business may deliberately pursue a new segment, but it should recognize the cost of doing so instead of mistaking every request for an obligation.
Translate observed pain into a business opportunity, not a requested feature
Analyze the process. Map the steps a customer takes, the tools used, and the time, money, or frustration attached to each step. The Pain and Frequency Matrix compares how painful a problem is with how often it occurs. A task done rarely can still matter when mistakes are costly; a frequent task can matter even when each instance is mildly annoying. Low-pain, infrequent tasks are less promising starting points.
Apply business judgment. Customer need is only one test. A solution also has to be understandable and usable, commercially sustainable, and possible for the company to build. When someone requests a feature, ask for a recent scenario, current workaround, frequency, cost, and people involved. Save the request and its context, look for repetition across customers, and revisit it if the evidence fits the company's direction.
The matrix is useful as a prioritization aid because it forces vague complaints into comparable dimensions. Its value lies in structuring questions, not mechanically ranking every opportunity. Pain and frequency do not capture every source of value: a rare task can be strategically important, and a frequent annoyance may be too cheap to solve profitably. The additional business tests prevent customer input from becoming a roadmap by referendum. A feature request is most useful as a clue to a job, not as a finished product specification.
Analysis
Deploy Empathy is a practical business guide built around a deceptively demanding premise: customer knowledge is gathered not by asking people to approve ideas, but by understanding the work and reasoning that precede their choices. Its narrative moves from the author's own transition away from guesswork to a usable method: recruit people with relevant experience, ask about actual behavior, listen without judgment, analyze patterns, and bring the evidence into decisions.
The book's most distinctive contribution is its treatment of interviewing as a learnable interpersonal craft. Silence, mirroring, and withholding advice are not decorative soft skills. They are ways to reduce the social pressure that can make participants perform, flatter, or defend themselves. The advice is especially useful for founders and engineers, who may have direct product knowledge but little training in eliciting another person's mental model.
The process lens is equally important. A product is rarely the whole task. Understanding adjacent steps, organizational roles, emotional stakes, and current substitutes can reveal why an apparently strong idea stalls. The healthcare product story makes this clear: technical usefulness and customer interest did not erase a long approval process that undermined the original commercial model.
The method has limits that the reader should keep in view. A small set of interviews can uncover mechanisms and suggest hypotheses, but cannot estimate market size or prove willingness to pay. Participants are self-selected, memories are imperfect, and a candid explanation is not the same as a purchase. Combining interviews with analytics, observed behavior, pricing tests, and inclusive recruitment makes the conclusions more robust. The book's strongest operational advice is to institutionalize curiosity without turning every interaction into research. Support, sales, and interviews have different obligations. When those boundaries are respected, listening becomes both a way to improve decisions and a way to treat customers as people rather than data points.
People Also Read
Glossary
Jobs to Be Done
Goal-centered product understandingA way to understand product choice by examining the broader goal a person is trying to accomplish and the circumstances surrounding it. A product is understood as one possible means to make progress, rather than as a bundle of features. Interviews explore the process, alternatives, motivations, and constraints behind choosing or leaving a solution.
Functional, social, and emotional
Three dimensions of a processA lens for exploring why someone follows a particular process or chooses a solution. Functional factors concern the practical task and constraints. Social factors include other people, expectations, and decision-makers. Emotional factors concern how the person feels about the task, the decision, and its outcome. The dimensions interact and can vary with context.
Research loops
Iterative narrowing of researchA cycle for exploring a broad or poorly understood problem. Interview an initial group, analyze what repeats and which needs seem promising, narrow the audience or question, and speak with another group. Repeating the cycle helps turn varied early observations into a more focused research question without assuming the first interviews provide a definitive answer.
Four Forces
Forces behind switching productsA Jobs to Be Done model for understanding why someone changes solutions or stays with an existing one. External pressures can push a person away from the current option, while the appeal of a new option pulls them toward it. Anxiety about switching and inertia around the familiar can hold them back. Interviews can reconstruct these competing influences.
Pain and Frequency Matrix
Ranks problems by pain and recurrenceA prioritization tool that plots customer problems according to how painful they are and how often they occur. Pain can include time, complexity, money, frustration, or the consequences of failure. Frequent problems may justify repeated payment; infrequent problems may still matter when they are costly or risky. The matrix identifies promising problems, but does not determine a solution or price.
Reaching for the Door question
An open-ended closing promptA question asked near the end of an interview to invite the participant to add anything important that has not yet surfaced. The interviewer thanks them, asks what else they think the interviewer should know, and leaves a pause rather than suggesting answers. It can prompt a final detail or concern, though some participants may have nothing more to add.
Switch interview
Reconstructs a recent changeAn interview with someone who recently adopted a product or another solution. It reconstructs the timeline from the original need through considering alternatives, deciding to change, and evaluating the new choice. Questions focus on past behavior, what triggered the search, who influenced or approved the decision, what concerns arose, and whether the solution is meeting expectations.
Valuable, usable, viable, and feasible
Four tests for product successA product evaluation framework that separates customer benefit from the other conditions for success. A solution must matter to customers and be usable by them, while also making commercial sense for the company and being possible to build. Customer interviews can illuminate needs and use, but the company must assess its own economics, capabilities, and constraints.
Download PDF
Download EPUB
.epub digital book format is ideal for reading ebooks on phones, tablets, and e-readers.