Raphael Thys
It doesn’t matter what industry you’re in, what your job is, or what outcomes you care about. Your success probably depends on your ability to influence human behavior.
But behavior change is hard — and we fail a lot. Think of all the apps, all the features, and all the startups that are created every day around the world. How many of those meaningfully impact behavior? Not many — which is why so many apps go unused and so many startups fail.
Design Thinking has helped us make progress by advancing the practice of empathic and innovative problem-solving, which is why it is considered the gold standard for product design. But Design Thinking does not provide methods or tools for identifying which solutions will change behavior before we build them. For this we need Behavioral Design.
Behavioral Design marries what the academic world knows about human behavior, and importantly, how to study human behavior, with what the design world knows about empathy and innovation. It builds on Design Thinking in 3 important ways: 1) It brings a more systematic approach to solution design, 2) It provides evidence-based theories for understanding user behavior, and 3) It offers quantitative research methods that enable teams to understand causality during the design process.
Behavioral Design
Behavioral Design is rapidly gaining traction in organizations around the world because it helps teams build more effective solutions with lower rates of failure. Teams develop a deeper understanding of the problems they want to solve so less time is spent on concepts that have no chance of working. Each solution becomes a datapoint that informs the next solution, which also accelerates future work. The result is more impact, faster.
Here is how it works.
Part 1: Systematic Product Design
In a traditional design process teams start by defining the problem they want to solve, and then they try to understand it better. They gather data and insights and then synthesize. They turn these insights into “How Might We” statements and then ideate. They use consensus to move ideas forward and then put winning concepts in front of users. They iterate based on the feedback, and eventually produce an experience.
This process is good, but it’s imprecise. It offers no tools or decision criteria for determining which ideas will have the greatest impact to behavior. So, decisions are made without enough information.
Behavioral Design brings more precision to this process. Through activities like behavior mapping, teams develop a more comprehensive and nuanced understanding of the problems they want to solve. They get very specific about the required behaviors a user must take to be successful, they map all of the steps in the process, and they identify all possible barriers along the way. This process is similar to journey mapping, but it is a bit more precise — and explicitly focused on behaviors.
As part of the mapping process, teams also carefully identify the psychological barriers that might not be obvious on the surface, but that can still make it hard for someone to take action. And because these psychological barriers come from the academic literature, teams can have higher certainty that they are meaningful barriers to behavior.
Once teams have a thorough understanding of the problem space, they systematically match solution strategies to barriers. These solution strategies are also psychological, and based in the academic literature, again leading to higher certainty around potential impact.
It is this problem-solution matching that creates strong hypotheses — and brings precision. Each proposed solution has a specific role to play in the goal of changing behavior — so the hypothesis is well-justified.
Problem-Solution Matching through Behavior Mapping
From this problem-solution matching, teams can create a “Theory of Change” to guide ideation, prototyping, and testing. The theory of change is like a recipe that explains the main ingredients of a solution — and importantly — how those ingredients work to influence behavior. This helps teams create a clear and logical articulation of how solutions should work, which gives them well-informed decision criteria as they build out concepts.
Behavioral Design Process
This approach brings a structure to insight generation and ideation that results in more targeted solutions and greater certainty around the potential for behavior change.
Part 2: Behavioral Science Theory
A systematic approach to product design is further enhanced by behavioral science theory. A good design process will always include research, data, and feedback from users. But not all insights give teams information about behavior change, because a lot of behavior is guided by psychological processes that are hard to identify and measure. This is where behavioral science excels.
Behavioral Science has made a lot of progress in understanding how humans operate — at a foundational level. The theories have been rigorously tested and replicated across many different populations and problems. When teams understand how humans operate, they learn to identify more meaningful patterns and find the signal through the noise.
How Humans Operate — Framework for Applying Behavioral Science Theory
This builds on what designers already do well, which is to make decisions based on as many insights as possible. Leveraging insights from behavioral science adds an additional layer of refinement and precision — which saves time in validation and creates a lot more certainty around hypotheses. And most importantly, it brings already-tested solutions to the table that teams might not otherwise know about.
Part #3: Quantitative Research Methods — throughout the design process
Design teams typically use quantitative data at the beginning and end of the design process. They set KPIs, they examine metrics and data science models related to their product or desired outcomes, and they take in all available customer research. And ideally when they launch a solution, they a/b test it and monitor their metrics as it goes live.
But during the design process, design teams rely heavily on qualitative UX research. As concepts and prototypes are developed, they’re put in front of users, and their input is used to refine the solutions.
Input from users is important, but it’s not enough if the goal is to change behavior. Users can tell you about their mental models and beliefs and we can observe a lot about their behavior and environments. But this will not tell us what will impact behavior, at least not with much certainty. That’s because so much of what influences behavior is driven by the complex processes in our brains that users are not directly aware of, and that we cannot directly observe.
Behavioral Design fills this gap by leveraging the quantitative methods developed in academia for the specific purpose of understanding and changing behavior. Because the standards for publication in academia are extremely high, these researchers have been incentivized for precision, conceptual clarity, and certainty around causal impact. As a result, the vast majority of Behavioral Science Theory comes from predictive surveys (that are psychometrically valid) and experiments.
Behavioral Design — Predictive Surveys
Behavioral Design Lab Experiment
With Behavioral Design, teams integrate predictive surveys & experiments into the design process. During behavior mapping, a team might run a survey measuring a number of conceptual barriers to see which are most strongly related to the behavior. This brings precision to the mapping process because the decisions around which barriers to solve for can be made with data. Teams can do a similar exercise with solution strategies, again making decisions with data.
Once teams have a theory of change and concepts, they can design lab experiments to test the causal impact of their solution strategies on intentions to behave, or other proxies for real behavior. Again, this brings more certainty and data-driven iteration on concepts before development resources are used building out experiences.
These quantitative methods can be easily implemented with the research tools organizations already use. Teams can use survey questions that have been validated by academic researchers and plug them into any survey platform (like Qualtrics or UserZoom). Then teams can run a regression with the survey data to create a simple predictive model that will tell them the strength of the relationship between the concepts that are measured and the outcomes they care about.
Most survey platforms also allow teams to build experiments. By randomizing survey participants to different prompts, framings, or even interactions, and then measuring intentions to take different behaviors, teams learn which concepts are stronger, at least directionally.
These methods do take some statistical know-how, but they are just as fast as typical qualitative UX research. Integrating these methods pushes teams to justify their hypotheses with data while they design — not just as the beginning or end of the process. This makes hypotheses stronger, which means the experiences that come out of them will have more impact.
Slowing down to speed up
Product teams often spend a lot of time debating ideas and how to implement them. And organizations build a lot of solutions that just don’t work. Behavioral Design brings a systematic approach, theory, and data to this process, which reduces time spent on decision-making, and reduces rates of failure.
The tradeoff is that Behavioral Design asks teams to spend more time up front in the problem space. But this tradeoff is well worth it because it produces less wasted effort in the long run. And over time, teams that apply Behavioral Design also become smarter. As they learn how to zero in on the most important drivers of behavior and update what they know with quantitative data, they gain better pattern recognition for the problem space as a whole. Each new solution builds on the solution before it, resulting in faster iterations and more impactful solutions.
So if you want to influence behavior, which you probably do, Behavioral Design is for you.