Survey Nonprofit Constituents at Your Peril

“Let’s start with a survey of participants about incentives to see what they want.” Two words should frighten you: “incentive” and “survey.”

In our work with recognition programs participants often say in surveys, “I wouldn’t want the gift I would earn for fundraising.” But when presented with a gift opportunity after earning one they redeem hand over fist. And higher fundraisers redeem at higher percentages than lower fundraisers, although the high fundraisers most often say, “I wouldn’t take the gift.”

What’s happening inside the average human head that doesn’t let them answer a survey accurately? The biggest problem is the kind of questions we ask. We ask people to predict their future behavior, and they don’t do that well. They can’t because one person is answering the survey (Present Self), and another person must keep the promise implied in survey answer (Future Self). Those two people have two entirely different sets of circumstances around them when it is time to answer versus perform, and they don’t behave the same way.

At conferences we hear phrases that made us itch: “Our fundraisers don’t think …” and “Our fundraisers don’t want …” and “The feedback I get is …” We interpret those phrases to mean, “We asked them, and this is what they said, so we took action based on their answers.” Unfortunately, we typically collect only Present-Self answers.

For example, I survey my child, “Would you mow the grass for $20?” My kid says, “I will mow the grass, dearest parent.” But after nightfall the grass remains uncut. My little survey failed to deliver accurate information.

My child’s Present Self answered the question to the best of his ability. “Not a terrible chore, and I need the money,” his Present Self thought. But his Future Self, who got signed up to mow the grass, did not buy in. Future Self thought, “I’m tired. I don’t want to take another shower today. I don’t want the money that badly.” Those are effectively two different people with two different decision-making data sets that led to two different outcomes. Or, another way to say it is, “I don’t feel that way anymore.”

Present Self is noble, looking forward to losing weight, curing cancer and feeding the hungry. Present Self is all about, “What I’m gonna do” and responds that way: “I will raise $5,000,” “I can recruit 10 people,” “I don’t need recognition,” “Of course I will participate again next year,” etc.

Future Self gets stuck with the bill. Future Self is the one that Present Self signed up for the walk, run or ride of choice. Future Self has a work project due, is suffering irritable bowel syndrome, has a hangover and was just recently victimized on some other front by Present Self. Future Self is angry, resentful, unappreciated and tired. Future Self says, “I’m bagging it.”

If we make decisions and plans based on those Present Self predictions of Future Self behaviors, we are most often surprised and disappointed when the surveyed group does not behave as expected.

Why is this sort of “would you …” surveying such a poor source of good information? Because, as Harvard psychologist Daniel Gilbert explains in his book, “Stumbling on Happiness,” it is easier to remember the past than it is to imagine the future.

The information resulting from Present Self surveys and anecdotal conversation is almost worthless. Highly skilled human resource professionals never ask an interviewee, “What would you do in this situation?” The best human resource people say, “Tell me about a time when this situation happened to you.”

The discussion around Present Self/Future Self is not new, nor is it hard to understand. What is hard is to have the discipline and insight to realize when we are collecting or relying on Present Self information. What is hard is evaluating what you currently collect to see if that is more worthy data. You may have sources of information in the form of registration reports, inbound call records, income records, redemption records, donor records, retention records and more. Surveys are more fun, but they yield less than the drudgery of data analysis.

Drop the “would.” If we ask people what they would do, what they would prefer, how they would behave, we will not get good information. We may get the answer we think we want, but it is not reliable information. Present-Self surveys are a failure at best and an effort at subversion at worst. This happens often at budget time, setting up revenue scenarios that are completely unrealistic.

The best use for Present Self surveys is when you are looking for a theory to test. An idea. Because that is all you are getting—a highly optimistic, noble thought is what most Present Self surveys produce. Potentially, Present Self surveys are dangerous to use at all as they can produce misleading information that will be interpreted as data, leading you down some costly avenues.