No Crystal Ball: Consumer Predictions in Surveys

In our surveys, we often ask respondents to predict the future. We don’t ask them to gaze into their crystal ball, but we ask them to provide a guess as to their future behavior. “This new phone will be released in six months. How likely are you to buy this new phone when it hits the market?”

In essence, we want people to tell us how much they will desire the new phone and how happy it will make them feel. Given the last blog about the limitations of memory recall, you probably know where I’m going with this. We, including the respondents in our research studies, are not particularly good at predicting our own future, much less how we will feel when it happens.

Fortunately and unfortunately, the future we imagine tends to be more extreme than the reality. We are likely to think that a negative event will be the end of the world, and we will feel the worst we have ever felt, and that a positive event will make us happy for ever and ever.

In most cases, neither is true. Not getting promoted will hurt, but probably not as much you think; you will live. You’ll be happy after buying that fancy car, but you will likely not be as intensely happy, and for a shorter period of time than you think. And in terms of how we will behave, we are motivated, biased, to see ourselves positively and to think good things will happen to us in the future.

Perhaps you’ve seen that commercial for life insurance, where the professor asks people to describe what happened in their lives in the recent past, and what they expect to happen in the future. The past was mixed, but people expected their futures to be very bright. (By the way, the guy in the commercial is Dan Gilbert, who is also instrumental on the work showing that people are bad predicting the emotions they will experience in the future).

So when you’re administering a survey, what should you do? If you’re asking people about their future, take it with a grain of salt. Recognize that what they are telling you is much more positive than it will actually be in reality. Make adjustments, calibrate the results. If it’s relevant, ask them about the past. The best predictor of future behavior is past behavior. What did they do the last time they were in this or a similar situation? Have they pre-ordered or waited in line for any new technology? If not, despite the fact that they really like your new idea for a phone, they probably won’t pre-order this phone either.

So there are limitations to what consumers can tell us. But there’s no need to throw the baby out with the bathwater. We are better off if we recognize consumer’s limits and adjust our expectations and the questions we ask accordingly.

And maybe design your survey so it’s easy to answer on that new smartphone.