We’ve been doing more user interviews here lately for various projects and investigations – for example for a Discovery Interfaces Review project, and to find out what students think of our study spaces and our response to noise reports in the library.
I also recently attended a workshop on Advanced Interview Techniques by Danielle Cooper from Ithaka S+R at the wonderful User Experience in Libraries – UXLibsV – conference.
Although I’m by no means an expert, I want to share some tips and tools you can use to conduct interviews yourself.
Why do interviews
Interviews are a great way to get qualitative insights from your users: what they do and why, how they feel and what they think.
Because the person is right in front of you, you can follow them if they go on an interesting tangent and you can ask follow up questions (like ‘What do you mean?’ or ‘Tell me more about that’) that you can never do in a survey.
In an interview, you’re not looking for anything quantitative (such as 4 out of the five people we talked to…). The responses you get will not be statistically significant. Instead, you will get rich and surprising detail and actionable insights.
What you need
- A script of relevant questions (see below).
- Users to talk to. The good news is, you don’t need very many (you will get a lot from 5 people).
- A suitable space: quiet, neutral, not too claustrophobic.
- A consent form (Sharepoint, IS staff only), because you’ll want to record them. Don’t try and take notes during the interview (except maybe the odd scribble); you won’t be able to interact properly with the person if you’re looking down and scribbling the whole time.
- A way to record the interview. On your mobile will do. I’ve used Voice Record for iPhone which lets me upload the recordings very easily into my DropBox. Next time I’ll try Otter Voice Notes (which is available for iOS and Android).
- An account on Otter Voice Notes: this does amazing automatic transcripts, which you can share with other people in your project team. You get 600 free minutes per month. The transcripts will be 80-90% accurate, but will need tidying up.
- Staff resource to tidy the transcripts and analyse the data.
What to ask
Be clear on what you want to find out about, and keep your questions relevant to that.
Ask about their experience, not their opinion or satisfaction:
- Do ask: Tell me about your experience using the library.
- Don’t ask: Do you like the library?
- Don’t ask: What do you like about the library?
Avoid:
- leading questions
- closed questions (where they can only give a yes or no answer), unless you want to establish a fact and then delve deeper into it
- vague questions.
- using our jargon.
During the interview
Top tips during the interview
- Be human.
- Keep questions to a minimum.
- Ask them to build on their answers.
- Ask for examples rather than for clarification.
- Encourage their narrative, not yours.
- Don’t interrupt, unless you’re running out of time or the conversation has gone off track (it’s impolite and ruins the auto-transcription in Otter.ai).
- Don’t justify our services (ie why something works the way it does).
- You don’t need to stick to the script – follow them along their tangents.
- You’re building a connection between this user and our department, so try and build a rapport.
To get more information out of people
- Use silence: they’ll want to fill it.
- Echo their response.
- Make affirmative neutral comments. (‘I see.’ ‘That’s interesting.’)
- Repeat or clarify the interview question.
- Summarize and synthesize the answer.
- ‘Tell me more.’
Managing interviewees
- If people talk too much: use gentle inattention, polite transitions, graceful interruptions.
- If people don’t talk enough: be sensitive, try a variety of probes; sometimes you just have to cut your losses.
After the interviews
- You’ll need time to absorb and analyse the data.
- Anonymise your participants’ responses, that way there’s no problem with GDPR – more on how to anonymise qualitative data.
- Remember that gathering data is only the start of the UX design process. Use the data to define the problem, ideate, prototype and test, before implementing a new or changed service.
More tips about user interviews: how, when and why to conduct them (Nielsen Norman Group)