Customer Discovery: Does personality matter?

01/06/13 21:27:00    

By Michael Mealling

Back in December David Cummings posted an article on The Art of the Customer Discovery Interview where he discussed how difficult they are to do and how not to “lead the witness” or introduce your own biases. As mentioned in the comments to David's article, examples for how to avoid these pitfalls were left as an exercise for the reader.

Amazingly enough, at almost the same time Pipefish was going through this exact process, but with the help of a professional interviewer. The interview technique we used is called “Laddering”. The idea is “where a seemingly simple response to a question is pushed by the interviewer in order to find subconscious motives.” Stealing liberally from Wikipedia:

It begins with a simple question, and then another question is asked about that response. For example, an interviewer may ask: “How come you skipped class?” and the response may be: “I went out with my friends”. The next question would be something like “Why did you go out with your friends?”. Essentially, the format is as follows:

Interviewer: “Why x?”
Subject: “Because z”
Interviewer: “Why z?”
Subject: “Because b”
Interviewer: “Why b?”

The first responses are generally functional justifications, like “I went out with my friends because I wanted some pizza”, or “I wanted some pizza because I used to eat it as a child”; but eventually the interviewer hopes to reach a virtue justification like “It's good to be childish”. Then it is fair to conclude that the interviewee skipped class because he valued childishness.

Those familiar with Agile software development will see a similarity to the 5 Whys technique for problem solving. It isn't sufficient to simply get a “thumbs up/down” response from a potential customer. A startup needs to understand the reasons why as much as possible. This is especially true with B2C oriented business models.

One of the things our interview coach helped us with is understanding the “personality types” (not the best word, but it works) of the users we were interviewing. You can introduce a severe bias by only interviewing people who are readily accessible. If your offices are on 5th Street near Georgia Tech then “getting out of the building” means leaving midtown because that area of midtown self selects for particular personality types and demographics.

The lean startup methodology is about being as intentional as possible. When it comes to customer discovery that means being intentional about who you are interviewing and how. Here are some examples for using laddering in interviews:

Should you hire someone to help you like we did? Its hard to say since I'm sure the costs are going to vary wildly. Customer testing and discovery is going to be a very iterative process for us so we're going to have to develop the ability to do that in house. There is a book coming out that might help so when it does I'll post a review.

Does anyone else have suggestions for user/customer interview techniques?


comments powered by Disqus