How to use this framework

There are lots of hard problems to solve in our world. We humans spend more money than we make, eat more calories than we expend, and worry about climate change from the comfort of our air conditioning. Despite good intentions and lots of interesting solutions, it’s just hard to change behavior. 

This framework is intended to be very general, but still useful. It comes from my years of experience doing behavioral research, designing products and experiences, and training people on how to apply behavioral science to every type of challenge you can imagine. What I’ve learned is that humans are humans everywhere — the same basic factors drive behavior, regardless of what the behavior looks like on the surface.

Understanding these 5 simple principles will help you think like a behavioral scientist:

1 — We all take mental shortcuts.
2 — Our emotions get the best of us.
3 — We’re egocentric.
4 — We’re social.
5 — So, context matters.

Putting it into action:

Step 1: Understand your problem, not just your user

Start with the behavior you want to create — your end behavioral outcome. Define it and then describe it. Next, think about how the principles may drive it. 

  • Is it something that happens when people are thinking deliberately? 

  • Does it say something about the individual? 

  • Does it come with specific emotions? 

Use the framework to really understand that desired behavior — and the things that get in the way.

Try to focus on the things that are true for most people who take the action — not just the things that are true of a few users you might talk to. And, most importantly, try to focus on the things your users can’t tell you directly — those subtle forces that the principles outline.

Step 2: Create a hypothesis about your problem

Ask yourself the following questions to generate a hypothesis:

  • What do you think is the biggest driver of the desired behavior? 

  • What do you think is most likely to get in the way? 

  • What could have a big influence on behavior, and importantly, why? 

  • Why would a successful intervention (product, experience, messaging campaign, etc.) lead to a behavior change?

Step 3: Design your solution to fit your problem

I call this “problem-solution matching”. This is the most important piece of the process — understanding your problem is great but it really only matters if it helps you design or apply the best possible solution. 

It can be tempting to get excited about a solution that worked well somewhere else. I have learned this the hard way! For example, you might see that communicating social norms gets people to use less energy in their homes. But that does not mean it will have the same effect on all challenges.

You have the greatest chance for impact if your solution comes directly from your hypothesis. You should be able to see — and explain — how your intervention brings to life the thing you think will create a change.

For example, if you think that people aren’t exercising enough because they experience discomfort when they workout, then communicating social norms around working out probably won’t help a lot. But, adding pleasure to exercise could minimize the discomfort and therefore have a bigger impact.

Step 4: Test whenever possible

We pay a lot of attention to what people tell us — what consumers want and like, what users think about experiences we design, or what people think would motivate them to shift their behaviors. But people can’t tell us everything we need to know because they can only tell us what they’re aware of or what they want to tell us. This framework provides the other piece of the puzzle — the factors that matter that are not obvious to individuals but become obvious in populations and with the right measurement tools. 

Thinking like a behavioral scientist can help, but doing research like a behavioral scientist is even better! Behavioral science could not exist if we didn’t have true randomized experiments. Experiments are the only way to measure impact with confidence and to mitigate/reduce the bias of individuals .

Experiments can come in many different shapes and sizes. You are probably familiar with A/B tests which can happen rapidly in digital experiences. On the other end you might also be familiar with clinical trials, which is how we test drug efficacy and safety. But there are also online “lab” experiments — often the go-to for academic researchers. These online lab experiments can be a great asset for designing products and experiments, in a scrappy and low-risk way. 

So get creative and try to test your interventions however you think you can!

More on why Behavioral Design is our Future here.