User Interview Tips

There are specific techniques you can use to retrieve actionable insights from an encounter with your users. How you conduct a research session will influence the quality of the data you collect.

You can increase your odds of collecting actionable data by taking painstaking care not to inject your own beliefs and assumptions into the discussion. This list is applicable to all data gathering with humans.

A. Make participants feel comfortable.

1. Respect their boundaries. Let them know they can end the session at any time. Let them know if there’s any question they’re not comfortable answering that they can pass on answering it. Adhere to the schedule for the appointment meticulously. Make sure they’re hydrated and physically at ease. Let them know the purpose of the session is to understand their beliefs, and there is no reason to fear giving a wrong answer as long as it accurately reflects their belief or attitude.
2. Don’t correct or judge them. Any correction or judgement will change the dynamic of the conversation and shut down the flow of necessary data from them to you.
3. Don’t explain things to them unless they’ve asked for an explanation, and then only after responding to the initial request with “what do you believe the right answer should be?” Remember you’re gathering critical data about user’s mental models here. Any explanation will change the dynamic of the conversation and shut down the flow of necessary data from them to you.

B. Listen actively

1. Probe for clarification when you don’t understand something they’ve said or done.
2. Do not make assumptions about their intent: ask them.

C. Avoid leading questions

Check all of your questions for hidden assumptions in the way they are framed. Questions that lead with “when did you stop…” or “don’t you think…”, or even “do you believe…” are encoding an implicit assumption that the user started or thought something. Questions which end with “you agree, right?” are forcing the user to evaluate and accept a conclusion they did not arrive at themselves.

D. Avoid dead end questions unless you need to move on to another research topic.

Dead end questions are generally answerable with a single fact. Yes, no, date, or number are appropriate answers for dead end questions, and when you ask them you’re not allowing the user to expound on their beliefs and assumptions. Only use them when you’ve gathered enough information on the current topic and you want to change the topic.

E. Avoid compound questions

The difficulty with compound question is knowing which part of the question the user responded to. “When you commute by bus, do you prefer to look at your phone or read a book?” is limiting the user’s options, even if the user has previously told you they commute by bus, have a phone, and read books. If your intent is to establish priority, a cleaner way to get that would be to ask “of all the activities you’ve said you do on the bus, which one is the most important to you?” Beware of questions that contain a pivotal “or” in the middle, as you’re limiting the possibilities and requiring an answer that is more complex.

F. Take copious notes.

Become adept at rapid note taking in a non-obtrusive way. Many researchers gain consent to create an audio recording so they can review the session later and codify their insights and observations. If you have a colleague willing to help, you could take a note taker along on the research session.