2.5 Discovering Customers: Interviews, Surveys, Observation
Customer Interviews
Interviewing customers is a powerful way to validate ideas. Before each interview, define the specific problem and identify your interviewees (e.g. “college commuters who use public transport”). During interviews, focus on the problem, not your solution. Ask open-ended questions that let customers tell stories. For example, ask “Who experiences this problem? How do you deal with it today?”. Avoid asking anything that suggests your product or pushes a certain answer. Let the conversation flow and probe deeper when something interesting comes up. For instance, saying “Tell me about a recent time you struggled with X” often yields richer insights than yes/no questions.
It’s also smart to summarize and confirm what you hear (“If I understand you correctly, …”). This ensures you really got their meaning and builds rapport. Near the end, ask, “What did I not ask that I should have?” – this often uncovers blind spots. Also ask for referrals: “Who else should I talk to?”. Finally, always thank interviewees and follow up; a little appreciation can even unlock more insights later.
Do’s (good practices) include: asking open-ended, unbiased questions about the customer’s past experiences (not future predictions), tailoring questions to the customer’s context, and listening more than talking. Don’ts (pitfalls) include: pitching your idea or demoing a prototype (don’t use interviews to sell!), asking for an NDA (NonDisclosureAgreement), or asking closed/leading questions. For example, never start with “What do you think of our product?” or “Would you pay $X for this?”. Instead, focus on real behaviors and pain. Remember: interviews are for learning, not presenting.
For detailed tips, see the YouTube video “How to do Customer Interviews” (Rob Fitzpatrick, author of The Mom Test), which covers these do’s and don’ts with examples.
Surveys
Surveys can reach more people, but only if designed carefully. First, keep surveys short and focused. Nielsen Norman Group advises only asking questions essential to your goals. Ideally, a survey should take only 5–10 minutes – any longer and respondents may drop out or give up. Use simple, everyday language to avoid confusion or bias. Avoid complex or leading phrasing (“we’re committed to 5 stars, how would you rate us?” is biased). Always write questions as if talking to a friend.
Choose the right question type: closed-ended (multiple choice) for quantitative data, open-ended if you need stories or opinions. Do provide balanced, mutually exclusive answer choices. For example, age ranges should cover all possibilities without overlap. Always include an “Other/None/Prefer not to answer” option for questions – this catches edge cases and prevents forcing people into bad fits. Don’t ask respondents to predict their own future behavior (“How likely are you to buy this in 2 years?”) – people are notoriously bad at that. Also avoid double-barreled or jargon-filled questions.
In summary, survey do’s and don’ts include:
Do: Keep it brief (ideally under 10 minutes), ask only what’s necessary, and use clear, neutral language.
Do: Logical order (start with easy, interesting questions) and group similar topics. Place any demographic questions at the end. Write for your target audience’s reading level.
Do: Give exhaustive answer choices; double-check numeric ranges and leave room for “Other/None”.
Don’t: Lead or bias the respondent with wording. Don’t require answers to every question (optional ones reduce drop-offs). Don’t ask about things you can find elsewhere (e.g. don’t ask age if you already have it).
Don’t: Try to get qualitative insights better suited to interviews (skip asking why questions that people may not answer in a survey).
Customer Shadowing and Observation
Sometimes the best research is unobtrusive observation. Rather than sit across from a customer, consider shadowing them in their actual environment (workplace, home, commute, etc.). This can be as simple as following one or two key customers through a day to see how they handle tasks. The goal is the same: discover problems they face without prompting. Many design courses emphasize this “fly on the wall” method. As one learning exercise notes, seeing customers in action “reveals new business opportunities” and makes problems crystal clear. For example, watching how a person cooks might reveal that they desperately need better kitchen organization.
Building Customer Personas
A customer persona is a semi-fictional profile of a target user. It brings your target segment to life by combining real data with imagination. Personas typically include:
Demographics (the “who”): age, gender, income, job, family status, etc. – basic facts about who the customer is.
Psychographics (the “why”): goals, values, frustrations, interests – why customers care, what motivates them.
Behavior (the “what”): patterns of use, buying habits, product preferences – what the customer actually does and uses.
For example, a persona might read: “Ella, age 22, a college student (demographics), who hates waiting in line and loves fitness (psychographics), shops for quick healthy meals on campus using a mobile app (behavior).” A well-crafted persona helps the team stay focused on a specific customer: you think in terms of “Would Ella want this?” at each step. (As Delve.ai notes, personas give a holistic view of users by grouping common goals and pain points.)
Early Adopters vs. Mainstream Customers
Startups often succeed by first winning over early adopters – the small group who are most eager for a solution, even if it’s imperfect. These users tolerate early bugs and value innovation. In Lean Startup terms, your Minimum Viable Product (MVP) should aim at them. Early adopters give feedback and proof of concept. In contrast, mainstream customers demand a polished “whole product”. As one product manager notes, “MVPs are right for early adopters, but once you reach mainstream customers they require full features and quality”. For example, the early iPod was just a portable player for enthusiasts, but later Apple offered iTunes, stores and support for mass appeal. In summary: begin with innovators and early adopters, then use what you learn to refine the product before chasing the mainstream.