The first question you want to ask your customers is usually not the question you want answered.
For some reason, our first instinct is to ask a question that is one or two or three steps removed from the actual information we want. I can’t tell you why this is, but I can tell you I’ve seen it every time I work with someone to get user research done.
Let me give you an example from earlier this week:
“What’s your job title?” [first question]
Coworker asks if we can run this survey question on a certain section of our website.
Me: “Sure we can”, I say, “but what are you going to do with those results?”
Coworker: “We want to know if that person might potentially be a decision maker or not.”
Me: “Then maybe we should ask that directly — are you the person who can makes decisions directly on buying this service?”
Coworker: “Oh. Well… we also think some people might be gathering information for the person who is the decision maker. It might be good to know the distinction, since we really don’t know what they came there to learn.”
Me: “And there’s your questions — what did you come here to learn today, with some multiple choice options around whether they’re a decision maker or a helper or just curious.”
“What did you come here to learn today?”; [better question]
The trick is to not think about composing the question first. That’s where our best intentions — to keep the question short, or to make the question super-inclusive, or to try to ‘sound professional’ — bite us. Instead, work backwards. What information do you wish you knew?
If you had the richest possible channel of interaction — say, a face-to-face friendly conversation, where it’s okay to ramble a bit and use hand gestures and verbal emphasis and facial cues — what would you say to try and get at that information? Why would you phrase it that way? Why would you use those words or pauses?
Once you’ve figured that out, then you can try to reduce fidelity and shorten it into something you could ask in a survey or during a customer call. But going the other direction is like trying to take a tiny photo and blow it up huge — you get something distorted and blurry. And when you start with distorted and blurry questions, you get distorted and blurry answers.