User interviews: how to find treasure by asking good questions

Published on
September 21, 2020
in
How Tos
Joe Roberson
Your users have some gold for you. You'll find it, but only if you prepare properly and ask the right questions. Here's how.
Treasure chest on a beach

This is a guide to running a good user interview. It follows on from our article about why you should do user interviews as part of your user research. It explains how to prepare for an interview and the types of questions to ask. 

How to prepare

User interviews require good preparation. If you don’t prepare then you won’t find out what you need. It would be like looking for treasure without a map, a compass, a spade or a good pair of boots.

It takes effort to prepare but it means you won’t be digging up a load of unhelpful sand or data. 

1. Get clear on what you want to know

First up, this means acknowledging there are things you don’t know. It’s OK to not know them, or to have assumptions (we all have them). 

“Just acknowledging that there are things you don’t know you don’t know puts you in a much stronger position vis-a-vis reality.” Erika Hall, Design Researcher

Then work out what you need to know. We call these research questions. They are different to the questions you will ask in the interview. They are more about documenting what learning you need to move forward with the problem you are trying to solve. Erika Hall writes beautifully about the need to dig in the right spot. If you don’t work out where to dig, you’re unlikely to find treasure.

Bad research question: How do we get people to use our service?

Good research question: How do people go about seeking help?

Good interview questions: How do you manage the problem? Can you tell me about the process you went through when deciding to contact the service?

Use this worksheet to get started.

2. Recruit your interviewees

This is like finding the right island to go treasure seeking on. 

Who are your users or potential users? Can you describe them? Be specific then seek them out. You only need five to start with!

You might find them among your existing cohorts of service users. Or you may need to go further and enlist support from other organisations. 

Work out the most direct route to reaching them. Directness is important. The more people you have to go through to reach them, the longer it will take and the less responses you will get.

Write something clear and compelling about who you are looking for, why and what you need. Ideally put this on the web: on a page, or an online form. Then ask people to visit it or show it to potential interviewees.

Always ask for consent. Adapt your normal consent form into an online form using a survey tool. You can also ask a consent question at the start of an interview if you are recording it.

3. Write a script

This is a bit like getting your digging boots on or your spade out of the shed. 

Whether it’s your first or fiftieth user interview, it always helps to have a script. A script ensures you cover off all the things you need to explain at interview start. It also helps you start the interview with more confidence. You don’t need to follow a script word for word but you do need it written down and in front of you when you start an interview.

Your script should include:

  • Appreciation for the interviewee
  • What the interview is about
  • Why you’re doing it
  • What you’ll do with what you learn
  • What level of confidentiality applies
  • How long the interview will last
  • How, because there is limited time, you might sometimes move on to the next question (while sidetracks can be revealing, this gives you permission to keep the interview moving and focused on what you need to learn).
  • A warm up to build rapport, relax the user, show them they are valued. This could be a question or something you share. 

4. Write your questions

Every question should be...

  • Plain English, no jargon. Use language that your users will understand.
  • Open not leading e.g. ‘how did you find…?’ rather than ‘how useful did you find…?’
  • Open not closed - yes/no questions don’t give you detail. You want detail!
  • Specific. This means honing in on the specific things you need to learn about.

Talk normally. This isn’t a survey. People talk more when a conversation feels natural. 

Sometimes you might get stuck or want to know more. When you need a prompt:

  • Ask ‘can you tell me more’
  • Ask ‘why?’ (sometimes three times)

Prioritise

Before the interview, work out your most important questions. They may not be the first you ask, but you need to know them. Then, if an interview doesn’t go to plan, you can skip the other questions and ask your most important ones before you run out of time.

5. What to ask

I. Ask about behaviour

Behaviour is more reliable than predictions or views and opinions.

“Ask people what they did, to share their experiences and behaviours, not describe what they do or want to do.” - Dan Sutch, CAST

Don’t ask about future behaviour. We are bad at predicting it. Asking about it will get you a hypothetical answer. It’s risky to make decisions based on these. 

So, avoid hypothetical questions like ‘would you use x…?’. Instead ask practical questions about things like:

  • Past behaviour
  • People’s lifestyle
  • Digital habits
  • How they’ve tried solving or managing problems in the past
  • How they experience or manage problems at the moment
  • Digital access: how they complete tasks online, and what makes it difficult for them

Giving timeframes will make it much easier for the interviewee to answer, like ‘in the last six months, how often have you used x?’.

Include tech or service habits

Knowing what technology your users are using is crucial to building any new product or service. It can also help debunk some of those long-held internal assumptions about your users. 

II. Ask about experiences

Ask about their good and bad experiences. Listen to how they felt, what they saw, what they did, what they used, how they experienced it. 

This will help you understand:

  • Their expectations of a good experience or service
  • What they need to make it good, or to avoid it being difficult. 

Remember to look beyond their experience of your service. You’re here to learn about their broader experiences of the problem, how they cope and how they behave.

“A good question is 'Tell me, when you did X or when X happened to you, what went through your head?'. Then people tend to describe their feelings and thoughts in a natural way that is really valuable to you.” - Ab Brightman, Catalyst

III. Treat views and opinions carefully, and ask 'why?'

Views and opinions aren’t so useful. They tend to be based on people’s ideas, thoughts and beliefs. It’s more useful to try to understand their motivations, the root causes of their experiences and their expectations of your service and the world. 

To find these out you need to ask ‘why?’. When you receive a view or opinion either enquire why, or move on to a more practical question.

This advice applies particularly to asking people what they want. Don’t just ask people what they want. Instead ask why they want it. That has much more value.

“There are some scenarios where I find asking 'what do you imagine X will be like?' can be handy for drawing out expectations. So you're not asking what they want, or for them to predict their behaviour but instead to paint you an image. E.g. asking 'what do you imagine would happen if you went to the sexual health clinic?' could reveal that they've been avoiding it because they think there will be a difficult conversation around being gay.”  - Ab Brightman

Need more help?

Shifting to user interviews and the curiosity mindset that goes with them is a process not an event. We’re here to help you make the shift. Use a Design Hop to get started, or even a Digital Candle call for some 1-1 support. 

And if you think this article could be useful to others, please share it.

Photo by Suzy Hazelwood from Pexels.


If you liked this article and want to learn more about how digital can help your charity, then sign up for our free weekly newsletter. Thank you!