How to conduct good user research and why it’s so important
Do you want to start doing user research, but don’t know how? Or have you been talking to users, but got disappointing results? If yes, you are not alone.
Talking to users is one of the highest-leverage activities an organisation can invest in, but is also difficult to get right.
This is a one-stop-shop introduction to the fundamental principles behind good user research.
Why it’s so important to get research right
You are not the user
We often assume, without realising it, that everyone thinks like we think and likes what we like. Generalising based on our own experience is actually a cognitive bias called the availability bias.
Designers and business leaders must fight this bias by interacting as much as possible with users.
People don’t tell the truth
But there is a catch. Talking to users does not automatically yield useful insights, and very often can lead in the wrong direction. Users don’t purposely lie to us, but things people say are different from things people actually do.
It takes skills to make them reveal their actual needs and behaviours. Without those skills, user research becomes useless at best, and misleading at worst.
Everything is MUCH easier when the problem is clear
If done right, User Research is a formidable foundation for everything that follows. Having a deep and reliable understanding of the users and their problems can empower a team like no other thing.
Decisions become easy because you don’t have to guess (no more design by committee). With this knowledge, you can create products orders-of-magnitude better, and save great amounts of time by avoiding rework.
User Research is not validation
Wait ✋. Before diving into the guide, here is an important disclaimer. User research is not “validation” or usability testing. User Research is mostly performed at the beginning of a project, and most-of-all, is agnostic of any solution. It is focused on deeply understanding the users.
An introductory guide to good user research
1. Define your Research Objectives
What is it exactly that you want to learn? Write down your research objectives before doing anything else. You should usually aim for 3–7 objectives per project.
Here is an example: “Understand what people need in order to feel prepared for their flight”. You could also try to learn about:
- Situations: the different kinds of situations that your target audience is often in, like running late to the airport or traveling with kids.
- Goals or Motivations: what are their goals with air travel?
- Knowledge: what is common knowledge about the air travel process?
- Decision-making process: how do they evaluate and select things like seat upgrades?
- Behaviours: what kinds of actions do they typically take and in what order?
2. Decide who you want to interview
You want to understand the actual users, but also their ecosystem. Interview users with a range of experiences and background. In some cases, you also want to interview non-users that have a supporting role.
For example, for the air travel example above, you would want to interview people that fly very often, people that have flown 1–2 times only, people who took short flights vs. long international flights, as well as staff.
If you were to do user research for an e-commerce website, you might want to interview people with different shopping habits (one-time buyer, regular buyers, company accounts, …), people using the competition, customer support, …
3. Design an interview guide.
How to structure your interviews. The first step is to look at your research objectives and brainstorm a list of questions that will help you accomplish them.
Then, it is time to write an interview guide. A good interview is structured like a funnel. It begins with a really wide general conversation and it becomes more and more specific as you proceed.
- Set the stage. Explain your goal to the interviewee and make them feel at ease.
- Prompt them to tell you stories, and ask follow-up questions to dig deeper.
- Finally, ask questions that help them summarise what’s important to them. Example: “If you had an advice for someone going through the same experience, what would you tell them?”
Ask for stories. The best way to get that information out of users is to get them to tell you stories. When your interviewees tell you stories about their experiences, they will also inevitably answer many of your questions all at once.
👍 Use open-ended questions: “How do you currently …?”, “What is it like to…?”, “Why do you feel…?”, “Can you tell me about a time when…?”.
👎 Avoid these types of questions: “Do you think it’s a good idea?”, “What should the product do?”, “Would you buy a product which did X?”.
Stay away from opinions. Never ask the user what they want. Designers are not the user, and the opposite is true as well: users are not designers. ️💡Note that feature requests can actually be a great start toward interesting learnings as long as you ask “why”. Why do they want you to add this particular feature? What are they trying to accomplish with it?
4. Conduct the interviews
Interviews typically last anywhere between 30 minutes and 1h30, depending on how much you can learn. When scheduling interviews, a good start is to default to 1-hour slots. Don’t hesitate to ask for another interview slot if you run out of time.
Here are some tips to get the best out of your interviews.
Always anchor things. Always try to come back to facts in your discussions. Anything else is worthless. What to look out for: generic claims (“I usually”, “I always”, “I never”), future-tense (“I would”, “I will”) and maybes (”I might”, “I could”). Whenever you spot them, look for what the users actually do, and why they do it.
Do not talk about your idea. You will skew your data by doing so, as people tend to twist their narrative to tell you what you want to hear.
Take as much notes as possible. Even better, have someone take notes while you lead the interview. You will most likely forget very quickly what you did not write down, so this is really important. Always clean up your notes right after your interview, as some of it might not make sense to you anymore later on.
When interviews go bad. Sometimes, you will feel like you are not learning anything from some interviews. This is the signal that you need to change something. It might be that you are not digging deep enough during the interview: did you ask “why” enough? Did you actually get to talk about concrete examples? It can also mean that you have covered the topic, in which case you should quickly move on to lower-level experimentations such as prototyping, or drop the topic if no opportunity emerged.
5. Summarise your learnings
After completing all your interviews, look for patterns. The next step is to create artefacts that will help your team build the best product or service, and spread your knowledge. Those can be personas, customer profiles, experience maps, strategy canvas, and so on.
How often should user research be done?
Unless you want to bet your whole product or project on intuitions, you should always start with some user research to ground your work in evidence.
As a rule of thumb, do a round of research every time you start something new, or you feel like you don’t understand a part of the problem at hand well enough.
In a past organization, each product team was doing a round of research (spanning 2–3 weeks) once per quarter, because we were exploring unknown areas. A good foundation of user research can last for many months if the product area stays the same.
☝️️ You should talk to users and customers as much as possible though, be it through research or other methods.
Further exploring
That’s it for the basics of good user research! Now it’s your turn to go talk to your users and collect great insights. There is much more to uncover on the topic, which is why I have gathered my go-to’s for user research below.
