Interviews

Interviewing people seems pretty straightforward, but interviews are a special type of conversation with formalities and challenges and tricks. The basic process for conducting interviews:

Make a list of questions; the more the better
Recruit interviewees
Conduct the interview
Review your notes

1. Make a list of questions

Before you begin scheduling interviews, write down the questions you’re trying to answer with your research. Since interviews are great ways to collect qualitative data, most of your questions should focus on the subjective and personal aspects of whatever you’re researching. It might be helpful to prepare a script for your interviews (but don’t read directly from it) that goes something like this:

Introductions

Remind interviewees who you are (or introduce yourself) and that you’ll be interviewing them today. Tell them what you’re interviewing them for, generally something vague like understanding how people think about _____, or developing an ad campaign for ______. Ask “Is now still a good time to chat?” to subtly make them interviewee feel accountable for the conversation. If you’re going to record the conversation (iPhones have a voice memo app), make sure you get their permission; tell them how you’ll use the recording, who you’ll share it with, and when/if you’ll destroy it. If you’re going to take pictures (and I recommend that you do), make sure you get their permission; the goal is generally to have one or two casual portraits and a few shots of the environment or objects that come up in your interview; tell them how you’ll use the photos, who you’ll share them with, and when/if you’ll destroy them.

Background

It’s good to ask a few basic questions at the beginning of the interview to get people comfortable (to forget they’re being recorded, to ease them into conversation, to get them answering quickly based on things that are generally easy for them to say).

Ask basic, factual questions relevant to your research. For example: How old are they? What do they do as a job? Where do they live? How many people are in their family? Where do they generally eat dinner? Who does most of the cooking? etc. Notice the questions go from really broad facts to more relevant facts to the research while still staying pretty simple. Notice the lack of Yes/No questions. Use those sparingly.

You generally want about 5-10 background questions.

Most of these questions should be covered in your screening (see the next section about recruiting interviewees).

Questions

After you’ve gotten some rapport with your interviewee, you can dig into the substance of your conversation. Generally, the more your interviewee talks and you listen, the better you’re doing. Gently bring them back on topic if needed. Make it a conversation, not an interrogation. You should feel like you’re asking a friend to teach you something, not confronting someone in court or hosting a talk show. Don’t judge anything. Just observe.

Ask the interviewee to take you through a process or activity from start to finish. Use the think-aloud protocol or have them act it out in the environment by moving from space to space. (For example, first I fill the sink with water. Then I get the vegetables from the refrigerator…. move to each of those places. Mime the action along with the interviewee.) Ask people to show you. Ask for stories.

Ask for comparisons. Best/worst. “Like that time when…” “What else is that like?” “What does that remind you of?” “How does that compare to…” Figure out “what’s the most important thing about…” or “what do you do first when…” or “when do you…”.

Look for gaps in what they’re telling you. For example: “You mentioned your husband sets the table while you’re cooking. What are your kids doing during that time?” And, look for exceptions. “You said you usually _____. What throws you off?” or “What else happens in a typical day that didn’t happen yesterday?” or “What interrupts you when you’re….”

Ask what frustrates people. Ask “what stops you from…” or “what’s holding you back from…”.

Never, ever assume you know an answer to a question you’re asking. Don’t offer possible answers. Don’t ask yes/no questions (except for basic background information). Strive to always ask questions that people can interpret in their own ways and come up with their own answers. Bad question: “Would you agree…?” Better question: “What do you think of…”

Always be looking for “why” someone does something. People are much better about explaining the past than predicting the future. Make distinctions. Ask for the definitions of important words (even of words that seem obvious to you). Ask lots of naive questions rather than trying to demonstrate your knowledge. Sometimes the simplest questions give the best insight into how someone thinks about things.

Follow the interviewee’s emotions by asking more questions whenever there’s an emotional reaction to something. (Be reasonable: you don’t want to make someone angry or have them cry, but you do want to push a bit further than might feel comfortable with a stranger.) Sprinkle in a few “How does _____ make you feel?” if you’re having trouble capturing emotions.

You generally want to have 20-30 substance questions written down, or about 4 questions per minute. You will go through questions much faster than you think.

Closing

Ask if there’s any question that they’d wish you’d asked them. Thank interviewees for their time and candor. Ask if there’s anyone else they think you should talk to about this. Thank them again. If you’re planning to share some sort of results with them, let them know how and when you’ll do that (this isn’t required).

2. Recruit interviewees

Generally, when we want to speak with someone, we try to capture people at the extremes: who is really obsessed with whatever you’re investigating? Who couldn’t care less? Who’s in the middle? The more you have at either end of the spectrum, the easier it will be to make compelling personas and inferences about the middle group of people. Plus, it’s much more fun to interview passionate people. You want a mix of people with a mix of backgrounds so that your data is somewhat representative.

If we’re doing a super formal study, we’d filter through lots of people using a “screener”, which is a fancy way of saying a flowchart of questions that are designed to disqualify people quickly. If you’re only interested in women who work and have more than two children and eat Thanksgiving meals with their family, you’d ask those questions in a simple Yes/No or multiple choice way to quickly find a few people who match your criteria. Generally, you want to ask the questions in order from least to most narrow/restrictive so if you’re having trouble finding the pefect matches for your criteria you can go back and talk to the almost-perfect matches again. You don’t need to make a screener or heavily filter the people you’re talking to for this project as long as your topic is general enough that it applies to everyone you’re interviewing.

Generally, after the fifth or so interview with a particular kind of person, patterns emerge and there are diminshing returns to interviewing more people (it costs more than it’s worth).

Most interviews (and research in general) involves some sort of compensation for the participant. In additon, a thank you note is always great.

3. Conduct the interview

Be punctual. This makes a huge difference in showing respect to the interviewee.

Dress to match what the interviewee will be wearing, roughly. Don’t wear a suit if they won’t be wearing a suit. Don’t wear workout clothes if they won’t be wearing workout clothes. But, don’t overdo it. You shouldn’t be buying clothes to conduct and interview, and you don’t want to be a poseur.

Interviewing people in context (at home, at work, at the gym, in the car, with their kids, while they’re relaxing, at the bar, etc.) is generally more useful to you than over the phone or in a sterile setting. (That’s one more reason to dress to fit in.) Document the people, objects, environments, messages/media, and services (“POEMS” as a mneumonic) that are involved in whatever you’re researching. If, for example, you’re researching Thanksgiving cooking, you’d want to take pictures of the refrigerator, grocery lists, where the kids hang out in the kitchen, who is involved in the preparation, where people sit, any coupons collected, any plans for the party (invitations, a calendar with tasks marked on it, a todo list, etc.).

Go slow. Leave long pauses for the interviewee to think. Don’t ask more than one question at a time. Leave some silences just to see what the interviewee will fill them with. Going slow feels very awkward at first, but often the most interesting things happen when you’re going slow.

Mirror the body language of the person you’re interviewing. If they fold their arms, fold your arms. If they slouch, slouch; if they sit up straight, sit up straight. If they smile, smile back. Etc. Mirroring is a way to make people feel more comfortable, to build empathy in yourself, and to show you’re actively participating in the conversation. Obviously, this can be taken too far and seem like mocking. Avoid that. Be suble. Use your judgment.

Take a buddy. Interviewing with two people is much, much better than doing it alone. One person generally leads the questions while the other person takes notes, pictures, notices emotions on the face of the interviewee, and chimes in when there’s something the interviewer might want to ask more questions about.

Avoid reading from a script. Feel free to take your list of questions as a reminder, but strive to do things from memory (naturally, like a normal conversation) and only glance at your outline throughout the conversation to make sure you’re not missing things.

4. Review your notes

After the interview, but before you go deep into the next task in your life (taking the train home, calling a friend, switching activities), take a few minutes to review your notes and jot down anything you missed. Mark what you think are the most important parts of your notes. Make a separate list of the top 5 things you learned in the interview. Then, revist your notes later to expand on them.