| | | |

Preparation is the key to a great research interview (Ask Dr. Wobs)

Photo: ARD

Preparation is the key to a successful research interview. Deliberate planning makes sure your time isn’t wasted — which is the concern expressed in today’s reader question.

Dear Dr. Wobs:

Good analysis requires research: analyzing data and interviewing people. I know how to make my research data sing. But how do I prepare for and conduct great interviews? Is ROAM analysis a good starting point?


Yasin, I’ve been thinking about this a lot lately. I’ve conducted over 500 research interviews in my work as an analyst and author. It’s become clear to me that many researchers haven’t yet mastered the art of effective interviews. (As with all the Ask Dr. Wobs questions I answer, I’m sending a free signed copy of Writing Without Bullshit to Yasin; if you have a good question, ask it here.)

Your suggestion to use ROAM analysis (Readers, Objective, Action, iMpression) isn’t quite right; those steps are for preparing to communicate to others in writing, while an interview is mostly about listening.

But preparing first is still the key. Lining up an interview with an busy person isn’t easy; it typically takes weeks, and you’ll be lucky to get 30 to 45 minutes. Having lined up the time, you want to make sure to use it efficiently. The biggest mistake is to get to the end of the interview, then start writing, and find out you forgot to gather some key piece of information. Subjects rarely have time for followup interviews, and followup emails are hit or miss.

So here’s how to prepare, and how to conduct the interview, and how to make it as effective as possible.

Set up the ground rules and the tech ahead of time

The worst mistake you can make is to stumble out of the gate. To avoid screwups, use email to set things up ahead of time. When you’ve set things up, put the following information into both an email and a meeting invitation:

  • Specify exactly the way you’ll connect telephone. In-person interviews are great, but you’ll conduct most of your interviews by phone. Share your phone number; ask for theirs. Will you call them, will they call you, or will you both call into a conference line? Or will you use Skype, and if so, will you be using video or just audio? (While there are many technologies similar to Skype, Skype is the one that most people belong to and are familiar with.)
  • Don’t screw up the time zones. Believe it or not, based on my experience, this is most likely reason you’ll miss a connection (and it makes you look like an idiot). Did you mean 2PM my time, or your time, and is one of us traveling and in a different time zone altogether? If you’re in New York, you get used to San Francisco being three hours behind and Paris six hours ahead, but even that changes during the Daylight Savings Time/Winter Time changeover. All invitations should include the time zone of all participants.
  • Share this info in a meeting invitation. By sending a meeting invitation, you get the info into people’s schedule, where it’s easy to see exactly as needed. And meeting invitations from Outlook, Google Calendar, or similar applications will automatically appear on both your schedule and your interviewee’s in the right time zone.
  • Get backup information. To prepare for a missed connection, request your contact’s mobile number or the phone number and email of an assistant that you can call if they’re late.
  • Who will be on the line? Will anyone other than you be listening in or asking a few questions? In my interviews, research assistants often listened, but it’s dishonest to include lurkers without telling the interviewee. And salespeople don’t belong on research interviews.
  • What are the ground rules? I typically share the following description:”This is an on the record interview. However, as we conduct the interview, if there is something you would like to share off the record, let me know at the time. Before I publish anything, I’ll run the information and quotes by you so you can check that the facts are correct.”

    This is a nuanced statement. It indicates that you’ll allow the interviewee to do fact checking, but does not promise them they can change or delete something that they shared, nor that you will make them look good. It’s far better to share this ahead of time than try and negotiate it afterwards, so get it into your emails and repeat it at the start of the interview. (It’s not the best way to warm things up, so deliver it in a quick and friendly way.) In some cases, people will request to be anonymous. In those cases, you can make a statement like this: “I’ll identify you as the CMO of a large packaged goods company but won’t share your name, OK?”

Research your interviewee ahead of time

You don’t want to waste time on stuff you can find on the Web. So find out who your interviewee is — where they worked before, how long they’ve been at the company and in the job, and what’s been written about them already. LinkedIn is an essential resource, but don’t stop there.

Your interview will get to the meaty stuff more quickly if you can ask questions that are more like “I saw in Ad Age that you had a highly effective email targeting program — can you tell me about it?” rather than “What did you do at Coca-Cola that was notable?” Such questions communicate to the interviewee that you’ve taken the time to research them, which shows respect.

You don’t have to do the research yourself; if you have a researcher or research assistant, you can have them do it. But that only works if they share the highlights of the research ahead of time, and you actually read it. A typical research brief for an interview is 200 to 400 words organized for quick scanning, not 5 pages of pasted-in background that the interviewer has to wade through.

Choose your best method for recording the interview content

There are two philosophies here.

I’m a very fast typist, so when I conduct an interview, I am creating in real-time an 80%-accurate transcript of the meaningful things that the interviewee is saying. My transcript is full of omissions and spelling mistakes, but it becomes an essential resource for later, and the act of typing helps me remember.

The other method is to record the interview, which is easy to do on Skype and many dial-in conferencing services. You can take much more fragmentary notes if you do that, but it’s time consuming to go back and listen to interviews over again. There are also transcription services that will turn the interview into text, but they’re not 100% accurate.

You can have an assistant take the notes, but I don’t count on them. I find that what I think is significant is rarely in exact congruence with what my assistant decides to take notes on. Notes from an assistant are more valuable as a backup and for training purposes than as a primary resource.

Prepare your questions with the end in mind

In any interview, there are things you must know. For example (and your list will be different):

  • What was your objective on this project?
  • What steps did you take, and in what order?
  • What quantitative results can your report, such as sales increases, cost savings, compliance improvements, or time savings?
  • How much did it cost? How long did it take?
  • What were the main obstacles, and how did you overcome them?
  • What did you learn that most people don’t know?

Prepare your questions ahead of time with these answers in mind. I often put the questions right into the document where I am taking notes, to remind me. You can also include things you’ve identified as interesting from your research, such as this being the first project to use SnapChat and get measurable results, or that the person said they were almost fired during the project.

Be realistic. You’re not going to get to ask 18 questions in a 40-minute interview. Pick the ten or twelve that matter most; recognize that you’ll probably only get to eight of those.

Arrange your interview as an hourglass

An hourglass starts broad, gets narrow in the middle, and the becomes broad again. Your questions should match this profile.

Start with some softball questions to warm up the interviewee. For example, “Have you always thought of yourself as a marketer?” or “What kept you working at the company for 15 years in a row?” These sorts of questions don’t always surface useful insights, but they warm up the interviewee and form a rapport.

Now narrow down. Start asking questions about projects and numbers and obstacles. These questions will require the interviewee to think harder and use judgment, but they’re more likely to do that if you’ve made them comfortable first. (Nobody starts a conversation with “So, how many millions of dollars did it cost to build that A.I. system, and why did it take four-and-a-half years?”)

At the end, get general again. For example, “What can marketers learn from your experience?” or “What’s next for you?”

Interrupt as necessary, and press for color

You are not an interview robot. Your job requires more than just listening. You must make sure to get what you came for.

Interviewees have a tendency to ramble and go on at length. Eventually, this uses up your interviewing time. If someone is talking, but not generating useful insight, you need to interrupt and get them back on track. For example, you could say, “Um, let me stop you there and see if we can focus specifically on the technology tools you used. I know my readers want to hear about that. Can you share more about that?” I recommend a tone that is polite but a little urgent. If they spend half the time talking about their mother’s immigration journey from Poland, you won’t get to what you’re looking for (unless your piece is actually about immigration journeys from Eastern Europe).

You must also be opportunistic. Listen for the details that will make a piece come alive, and probe on those. The key member or the team was a musician — why was that important? You mentioned your boss getting in the way — what was the problem, and how did you get past it? You said you increased sales by 4%, but what were they originally, and how long did that increase take? Did you really start by testing your video compression algorithm on porn? You’ll find that interviewees want to share this stuff, with a little encouragement.

If you hear something that sounds great, say so. Your enthusiasm will encourage the interviewee. “Did you really code the app in just two weeks? How is that possible? What’s your secret?”

In business books, you should ask some questions that are more personal, like “What makes this project special for you?” or “How did you develop a taste for this sort of work?” People’s personal stories humanize them, and business books are full of stories about humans. If you’re creating a research report, on the other hand, these sorts of details are extraneous.

You also want to ask the question you think they probably won’t answer. “Weren’t you scared Microsoft would enter your market?” “How did you get past the FDA compliance regulations?” “What did you do after the project manager went to your competitor?” Best case, you’ll find out something no one else knows; I’ve discovered many interesting facts by asking these probably-won’t-answer type questions. Worst cast, you’ll piss off the interviewee, though, so ask these questions last, just before the general wrapup questions, in case they make the interviewee give up on you.

If you do this right, you’ll be able to write your interviewee’s story well

Your objective is to write what happened. This means a narrative with a beginning, middle, and end, with some statistics and color along the way. You might write up that story in a 2,000-word case study, or it might just be three sentences in a bullet point. But you’re only going to succeed if you’ve prepared properly for the interview. Then you can spend your time writing, rather than lamenting the questions you forgot to ask and the key information that’s missing from your case study.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.