Asking the Right Questions

by Aaron Shields
About this author

I’m struck by how often companies conduct research without asking any questions. The research efforts gather data but fail to give results beyond numerical characteristics (e.g., sales and demographics).

They expect that somehow from the mountain of data an answer will suddenly emerge in an eureka moment. But, the problem is that numbers don’t reveal answers.  You need to have a question.

You can start with a mountain of data, ask a question, and then pull out your mental shovel and dig through to find the hidden treasure. But how inefficient is this compared to having a question and then tailoring the research to match?

Asking ex post facto research questions also limits the number of questions you can ask, and, therefore, decreases the quality of the outcome—it forces you to fit the question to the research.

I would guess that this is one of the reasons why demographic research is popular—demographics are easy to accumulate without asking any real questions as they’re based on numerical census-style, generic questions, and they generate obvious answers: what groups are we serving and what groups aren’t we serving. The inevitable result is that we need more of some age and economic group. This is a poor question.

As Einstein said, if he had an hour to save the planet, he’d spend 59 minutes coming up with the question that needs to be answered, and 1 minute solving it. Marketers I’d guess would spend 1 minute coming up with the question and then 59 minutes solving it, only to determine that it requires annihilating all males in the 30-40 age group as they’re polluting the planet the most.

What do demographics really tell you that is significant? Do they just provide poor answers to poor questions?

I was sitting on a bench at Rittenhouse Square in Philadelphia when a man approached me and asked if he could ask me a few questions. He showed me the trailer to National Treasure 2: Book of Secrets, and asked me questions concerning my age, what I thought of the actors and director involved, would I see it in the theater, and what age group I thought would be most likely to see the movie. I told him I had no intention of seeing it in the theater but that I thought the most likely group to see the film was my age group. He looked puzzled:  how could I not want to see the movie but think it’s most appropriate for my age group?

Think about it for a second: A friend asks you whether or not you think he would like Juno. Would you say to yourself, well, he’s between the ages of 30 and 40, male, earns $50,000 a year, Hispanic, single, heterosexual, and lives in the Midwest, so I don’t think he’d like Juno. Or, would you think about the type of humor he likes, what other movies you know he likes and make your recommendation based on the way you see him as a person, instead of numbers?

If you didn’t bother to collect the proper information to solve the real problem, you’ll inevitably end up with a bad answer, unless by chance you happen to capture what you need. And the only way to be sure to collect the proper information is to start with great questions.

So, next time before you start gathering data, ask yourself what you really want from the data and plan accordingly.