How To Run Your First Customer Interviews

(Thanks for reading, I’ve recently written a more detailed version of this article here)

You’re probably reading this because I told you that you need to run some customer interviews, and it was late in the day so you agreed without really thinking about it, and now you’re working out what to do.
This is the perfect guide for you.

There’s an overwhelming amount of information available about running customer interviews, so much so that it’s hard to digest.
We’ll look at the basic principles and “rules of thumb”, with links to more sophisticated resources if you’re interested.

Rule 1: Interviews are not ads
The purpose of an interview is to understand your customer in more detail, not to persuade them that you’re a genius entrepreneur with a stunning idea.
In fact, your product/service might not even come up in the conversation.
We want to understand how customers think and act – how they evaluate options and how they measure if something is “good value”.
The aim is not to educate them or change their minds, but to understand how they think today.

Rule 2: You want to talk to 5-7 honest people
The danger of talking to fewer than three people is the low sample size distorts your data.
One cranky person looks like 33% of the market are cranky people, which may or may not be true.
Honesty is crucial, because if your participants start changing the truth, you might build the wrong thing.
Steve Blank said: “Cheating on your customer discovery interviews is like cheating in your parachute packing class”.
A great way to get honesty is to ask about past behaviours, rather than opinions.
Opinions are easier to misremember, whereas stories of their past behaviour are a bit more objective.

Rule 3: Open ended questions let people tell you the real story
Closed ended questions (Yes/No) are dangerous because they miss any sense of context.
To get a better picture, we want to give the interviewee space to tell us all the information they think is relevant.
You might not think it’s relevant, but we’re not interviewing you.
Open ended questions start with “Tell me about a time when…”, and allow people to describe a full scenario, with rich detail that leads to “aha” moments.

Rule 4: The magic question
This simple question can squeeze out so much extra gold from your participant:
“and what else?”
What’s great is that it takes people out of autopilot – their brains are happy to give you their first instinctive answer, which is valuable but often incomplete.
Tacking on an “and what else?” to your questions might only work 40% of the time, but those extra details are often the most interesting.

Rule 5: All assumptions are guesses until validated
You’ve probably drawn out customer personas already, and they’re a good start.
However, your first guess might not be nuanced enough.
What if there are actually two different subsets of customers?
What if you’ve categorised people by what they buy, instead of why they buy?
What if you guessed the right cluster of Value Propositions, but not in the right order of importance?

Just because you guessed it in a workshop doesn’t make it right.
All of our assumptions are to be labelled as assumptions until we have evidence (not gut feeling) that they’re correct.
As Richard Feynmann pointed out: “It doesn’t matter how beautiful your theory is, it doesn’t matter how smart you are; if it doesn’t agree with experiments, it’s wrong.”

Rule 6: Find a second pair of ears
It’s really hard to conduct an interview while observing everything that’s happening.
You might be focusing on what the interviewee is saying, whilst also formulating your next question – this takes up all of your brain power.
Having a second person there is invaluable, because they can observe how the interviewee processes the question, their body language, and will pick up on what isn’t said.
You can debrief after the interview, and your differing observations will prompt some interesting points.
Failing that, you can use an audio recorder and a camera to record the interview (with permission).
I personally find that my phone does a good job in a small tripod on the desk, which gives you the ability to re-watch the interview later on.

Rule 7: Small incentives yield the best results
The interviewee is doing you a favour, so it’s polite to offer them something in return.
However, if we make the incentive too large we attract the wrong type of person, or bias their answers towards what they think we want to hear.
For example, I recently held some interviews with cruise ship passengers in Papua New Guinea.
We held a 90 minute “speed dating” session, where our accelerator cohort conducted short interviews and rotated around the room.
Each participant received $30 of onboard credit, and that’s the right amount.
It was enough to compensate someone for their time, but not lucrative enough to tempt people who weren’t our real customers.

Rule 8: Separate questions about their behaviour from your product
Customer empathy interviews are all about the interviewee, whereas customer tests are about the interviewee engaging with a prototype of your product/service.
Make sure you know which one you’re doing, because it will affect what you ask and how you structure your time.
You can do both in the same session, but just be aware that people’s responses will change once they know what your business is creating.

As an example, I was recently in an interview for Marvel Stadium.
The first half of the questions were about how the interviewee spends their leisure time – how often they go to restaurants, cinemas, sporting events, concerts, etc.
The second half involves the interviewee being shown a series of concepts for upgrading the stadium, then the interviewer asks them what they think of these ideas.
e.g. six ideas are laid out on the table, then the participant names which ones stand out as being of particular interest.
This is a great way to rank Value Propositions, but only after you’ve gathered unbiased data about how people have behaved in the past.

So to recap:
If you set up 5-7 conversations, focus on them instead of you, ask open ended questions, bring a buddy or a camera, offer a small reward and structure the questions in the right order – you’ll dig up some gold nuggets.
If you have limited time, I’d rather you talk to two genuine customers rather than eight of your friends and family or forty random people on a Facebook poll.

If you find this subject interesting, then you’ll love all of the great resources available online.
I recommend articles like How To Conduct User Interviews, the d.school Empathy Fieldguide, or Envato’s Techniques For Empathy Interviews In Design Thinking.
The book Value Proposition Design is great, full of ideas for tests and creative interviews.