10 Books That Will Make You a Better User Researcher

Most people who run user research were never trained to do it. They inherited it, usually alongside a design or product role, and learned it session by session. These ten books cover the four things that actually separate useful research from wasted research: running it properly, asking well, understanding why answers mislead, and getting a finding acted on.

Which books help most when you are running research alone?

Just Enough Research — Erika Hall (A Book Apart, 2013), is the right first book for almost everyone. It is short, opinionated and unusually honest about the conditions research is normally done in: not enough time, not enough budget, and a team that has already half decided what it wants to build. Hall's argument is that a small amount of properly framed research beats an ambitious study that never happens, and the book spends its length on how to frame a question well enough that the answer changes a decision.

Observing the User Experience: A Practitioner's Guide to User Research — Elizabeth Goodman, Mike Kuniavsky and Andrea Moed (Morgan Kaufmann, second edition 2012), is the reference to keep beside it. It is a broad practitioner survey covering interviews, diary studies, contextual enquiry, surveys, usability testing and analysis, with enough procedural detail to actually run each one. Nobody reads it cover to cover. It answers the question of how a specific method is really conducted, which is precisely the question that comes up when a study is already scheduled.

Rocket Surgery Made Easy: The Do-It-Yourself Guide to Finding and Fixing Usability Problems — Steve Krug (New Riders, 2009), narrows to one thing and does it well: small, frequent, do-it-yourself usability testing. Krug's case is that a handful of modest sessions run every month finds more real problems than one large study run once a year, and he explains the whole mechanic plainly enough that a first-time facilitator can follow it.

What separates a good interview from a pleasant conversation?

Interviewing Users: How to Uncover Compelling Insights — Steve Portigal (Rosenfeld Media, 2013, with a second edition in 2023), is the standard text on research interviewing and deserves that position. Portigal treats the interview as a skilled practice rather than a chat with a script: how to build rapport quickly, how to follow an unexpected thread without losing the study, how to sit through a silence instead of filling it, and how to notice the moment a participant starts performing the answer they think is wanted.

The Mom Test — Rob Fitzpatrick (self-published, 2013), comes from startup customer development rather than design research, and it is on this list for one specific contribution. It shows exactly how a well-meaning question invites a flattering answer, and how to rebuild that question around what the person has already done rather than what they say they would do. Anyone who has heard an enthusiastic "yes, I would definitely use that" and watched nothing happen afterwards will recognise the problem immediately.

Practical Empathy: For Collaboration and Creativity in Your Work — Indi Young (Rosenfeld Media, 2015), goes after the interpretive half. Young distinguishes between what a person did and the reasoning behind it, and argues that the reasoning is the durable, transferable finding. Behaviour changes when circumstances change. The thinking underneath tends to persist.

That distinction has a direct practical effect on what a researcher writes down. A note recording that someone rang the helpline twice before submitting a claim is a fact about one service on one day, and it expires when the service changes. A note recording that they rang because they could not tell whether an unacknowledged submission had failed or was simply slow describes a reasoning pattern that will reappear in the next service, and the one after that. The second kind of finding is the one worth the effort of collecting properly.

Why do people say one thing and do another?

This is not dishonesty, and treating it as dishonesty produces bad research. Thinking, Fast and Slow — Daniel Kahneman (Farrar, Straus and Giroux, 2011), sets out the two-system account of judgement that explains most of it: fast, automatic, associative thinking running most of the time, with slower deliberate reasoning invoked far less often than people believe. A participant explaining why they abandoned a form is usually constructing a plausible account after the fact, not retrieving a record.

Predictably Irrational — Dan Ariely (HarperCollins, 2008), covers similar ground more concretely, working through anchoring, the pull of free options, the influence of expectation on experience and the way context changes what looks like a reasonable choice. It is the easier read of the two and useful for a different reason: it gives a researcher vocabulary for explaining to a product team why a stated preference in a survey is weak evidence.

Neither book is a research manual, and neither should be treated as one. What they change is the weight a researcher puts on self-report, and how carefully a study is designed to observe behaviour rather than only collect explanations of it.

What to read when the finding is right and nothing happens

A great deal of research fails after the analysis. Storytelling for User Experience: Crafting Stories for Better Design — Whitney Quesenbery and Kevin Brooks (Rosenfeld Media, 2010), addresses that directly. It treats narrative as a working research instrument: how stories are gathered during sessions, how they are selected and shaped honestly, and how a well-chosen account of one person's experience travels through an organisation in a way a summary table never does. The ethical line matters here, and the book holds it. A story used well illustrates a finding; it does not stand in for one.

Quantifying the User Experience: Practical Statistics for User Research — Jeff Sauro and James R. Lewis (Morgan Kaufmann, 2012), covers the other half of the credibility problem. It is a practical statistics book written for user researchers, dealing with sample sizes, confidence intervals, task completion and comparison between designs. Qualitative work carries most of the explanatory weight in this field, and it is often dismissed in rooms where only numbers are treated as evidence.

Knowing what a small sample can and cannot support is what lets a researcher defend a qualitative finding without overstating it. It also makes it much harder for anyone else to overstate a number.

What none of these ten will do for you

A reading list is worth being honest about. These books will not tell you how research is commissioned, staffed and repeated inside a particular organisation, and that is often the real constraint. They will not resolve who owns a research repository, how consent is recorded, how participants are recruited fairly, or what happens when a study reaches a conclusion the roadmap has already ruled out. Those are local problems, and they are generally solved by talking to people in the organisation rather than by reading.

They also assume a fairly conventional research setting. Work with participants who are unwell, distressed, digitally excluded or dependent on the service being studied asks for care that a general practitioner text cannot supply in the abstract. Krug and Hall are candid about method under pressure, and Young is the closest thing here to a treatment of listening well, but none of them substitutes for domain-specific guidance and appropriate ethical review.

None of this is an argument against the list. It is an argument for reading it as a foundation rather than a complete education.

How to read the ten

Start with Hall, then Krug, and keep Goodman, Kuniavsky and Moed as the reference you open when a method is imminent. Add Portigal as soon as interviewing becomes a regular part of the job, with Fitzpatrick alongside it as a corrective and Young after that for interpretation. Read Kahneman and Ariely when you start noticing the gap between what participants say and what they do, and read Quesenbery and Brooks together with Sauro and Lewis when good findings keep failing to land.

The ten are not a syllabus to complete. They are four capabilities, and most researchers are noticeably stronger in some than others. The useful question is not which of these is best, but which one addresses the part of the work that is currently going wrong.

Companion resource: 📚 The User Research Reading List