Changing Behavior

We’re stubborn creatures. Our behaviors – good and bad – are hard to change. This can be frustrating for those of us in the business of offering people new products and services; “if they would just use this, their lives would be so much better!”

I’ve frequently equated design to behavior change. A common way I describe interaction design to people is that it’s the design of and for human behavior. There are a few lenses from which I view the word behavior:

  1. Things a person does at any given moment – a localized behavior
  2. Things a person does over time – a habitualized behavior
  3. Things we all do – a cultural behavior

Things a person does at any given moment – a localized behavior

When I was in school studying interaction design, many of our projects focused on getting people to do things in a way that we wanted. One of the projects that stands out was focused on physical behavior: how can we get a person to physically move in the way we want?

The constraints of the assignment were simple. Using a 4″ square block of pink, shapeable foam, persuade (ok, manipulate) someone to interact with it in a certain way. Get them to pet it, squeeze it, throw it, shake it, or rip it.

How would you get someone to pet the square? You could add a covering material that is soft, like fur or feathers. You could cut the foam square and reassemble it into a long, thin shape, and then sand that to be smooth – something begging to be stroked. You could create a sphere, embedded in a larger cavity like a track ball, so “petting” it moves the ball back and forth.

There are infinite ways to get someone to crush, rotate, pinch, or otherwise interact with an object. The focus of the assignment is on persuasive interaction – on identifying the way to get someone to behave in a certain way at a very localized level of detail.

This design for localized behavior is something a product designer does on a daily basis. It’s our bread and butter. We design interactions for both physical and digital products, and these interactions urge someone to behave in a certain way. Click this. Press that. Don’t go over there. Do these things in that order.

The design work for this is tablestakes, and is often grounded in the fundamental principles and processes of usability engineering. To design for localized behavior, designers need to understand basic principles of cognitive psychology, perception, motor skills, and memory; and, they need to understand typical design conventions and patterns, and expectations people have about how interfaces work. Designing for localized behavior is rarely about being innovative – it’s often about being practical or appropriate. If someone’s already figured out how to get people to pet a square of pink foam, it seems like a good idea to copy their precedent.

We see localized behavior design in, well, everything. The “buttons” on the screen on your phone aren’t buttons at all, but they sure look that way; the interface elements have shadows, line weight, sizes, colors, and proximity that say “tap”, “tap and hold”, “tap and scroll”, and so-on. They are the digital equivalent of fur on a square piece of foam. And wayfinding signage, physical design of walls and spaces, handles on doors and light switches are all examples of localized behavior design: designing for interactions.

Because this form of behavior design is so well established, we have a pretty good ability to predict if our designs are effective. We can compare them to best practices and heuristics, can evaluate them based on concrete measures (like button sizes or accessibility color ratings), and can conduct simple usability testing on our creations to understand if people use the things we’ve made “correctly.”

If we get this wrong, people may encounter complexity, feel confused, or feel dumb. They may make errors, they may be less efficient, and they may get angry. Our work has immediate impact that is material and measurable.


Things a person does over time – a habitualized behavior

Habitualized behavior happens over time, and with regularity. When we design for this form of behavior, our goal is to fit our product or service into the fabric of someone’s life.

My wife has a routine when she wakes up. She turns the water on in a water kettle, grinds coffee for a French press, and then enjoys some quiet meditation. When she’s done, she pours the water, now boiled, in the French press, feeds the cats, and explores the news online. After about 5 minutes, she “smushes” the French press and has her coffee. She finds this process calming, and it’s habitualized – it happens every morning.

If your company makes a new type of coffee maker and wants my wife to buy it, you’ll need to compete not just on price, features, availability, and so-on; you’ll also need to compete with her habit. It’s not that there aren’t alternative ways for her to start her day – there are countless ways for her to accomplish her meditation, news surfing, cat feeding, and coffee making. But she’ll need to see and value the new process more than her current one, and the emotional switching cost is high.

The design work for this requires an understanding of a few things about people. First, it means understanding what people currently do, but more importantly, understanding why they do it. People do things for all sorts of reasons. Sometimes, they have a job they need to get done. But often, our actions aren’t utilitarian; instead, they are tied to our sense of self. I may want to look or feel differently about my body, so I go to the gym. I may want to be around people like me, so I hang out at a certain type of bar. I could buy a specific brand to demonstrate or exercise my view of sustainability, or follow a sports-team to be a part of something exciting, or go to church to feel connected to a community. Or, I may make coffee with a French press because it gives me a period of waiting time that encourages meditation. “Why I do something” can be hard or impossible for someone to articulate, and when dealing with habitualized behavior, it’s rarely causal; it’s unlikely that when something discrete happens, I do something in response over and over. Instead, actions are complex.

In addition to understanding what people do and why they do it, designing to support or encourage habitualized behavior means understanding someone’s actual appetite for behavior change. I may want to quit smoking, but the actions I’ll take to do it are limited – my commitment to change is small. Often, this behavior change commitment is shaped by the culturally accepted norms around me. Google Glass overly challenged our norms around computing and privacy, but also around aesthetics: it looked weird. Segway was similar: it may have been a positive force for behavior change, but it made riders look stupid. These were tools that could have changed our behavior in meaningful ways, but we didn’t commit to them because they didn’t fit into our sense of self.

It’s interesting to juxtapose the failure of Segway with the success of public scooters, like Lime Scooter. Arguably, scooters don’t make people look their best, and there’s a long list of mockery of the items; as Wired describes, “it’s impossible to look cool riding one.” [www] But the scooters fit into our lives, because they emerged with a service wrapper that was friction-free. My threshold for behavior change is multi-faceted. The value of a just-in-time ride, the seamlessness of the service, the cost, and my ability to look cool – or dumb – are all in tension with and against one-another. To change my habitualized behavior around “how I go to work” requires a perfect storm of these qualities hitting all at once.

Designing to change habitualized behavior is hard. It’s hard because of the fuzzy and highly emotional background for people’s actions, because of how ingrained behaviors are, and because of how reluctant we are to try new things, particularly with technology.

If customers try to adopt our products, but we get this wrong, people may not change their behavior – our product and services may fail. This often manifests as our inability to get over the chasm Geoffrey Moore describes between early adopters and the early majority. Early adopters may buy a new product or service and use it once or twice, and feel as though they received value for their purchase because they got to be part of experiencing something new. But an early majority purchaser won’t put up with a product that they don’t feel succeeded. If I bought a tool that promised to help me lose weight, and I don’t actually lose the weight, I’m going to be disappointed. The fault may be mine, because it’s silly to expect habitualized behavior to change overnight. But as a customer, I’m certainly not going to blame myself: I’m going to blame the product I bought. And the market never blames the customer- it always blames the company.


Things we all do – cultural behavior

A cultural behavior is something that we’ve collectively established as “the way things are.” Some of these shared feelings and interactions are legally mandated, like driving on a certain side of the road. Some are emergent and constantly negotiated, like how to interact with a phone in a public setting. And some are primarily driven by technological leaps, as how our dependency on GPS has changed the way we think about where we physically are. While designers often focus on localized behavior in most product and service design, and habitualized behavior when we have the luxury of digging in deeper to a broader experience story (often from the shared perspectives of strategic as well as tactical design), it’s rare that we ever get the ability to explicitly design for a cultural behavior.

How do these things cultural behaviors change?

Watching the buzz and reality of self-driving vehicles may give us a clue.

People are enamored with autonomous vehicles because they bring to life a robot fantasy world: cars, moving without our assistance. But it’s the implications of the autonomy that are more interesting than the wheel turning itself. If a car can drive itself, it can come get me when I need it. This means it can be anywhere, rather than where I parked it. It can be blocked in by other cars, as long as those other cars can get out of the way when I want them to. If it can be anywhere, it could also be any car: it doesn’t need to be mine, it just needs to be where I want it, when I want it. And if I don’t need to own a car but can get one any time I want, then I could also get any kind of car I want. I could change to an upscale car for the fancy dinner, or a tough truck for the trip to Home Depot.

Consider the cultural behaviors that can change, and will have to change, for this to work – and how we’ve already inched our way there. We’ll need to get comfortable with getting cars on demand. We’ve already been trained that this works by hailing taxis or calling Uber. We’ll need to be OK with not owning a car, but sharing it. Car sharing programs like Car2Go have helped us see the value of this direction. Those of us who use Uber have become accustomed to hailing an Uber, and then waiting for several minutes. We’re comfortable with driving in a stranger’s car. And we put up with the rare occurrence of a driver who gets lost or cancels.

But many people have a much more established set of expectations around driving. They aren’t used to waiting a few minutes – they open their garage and drive away immediately. They don’t get lost, because they know the neighborhoods. And they find pride in car ownership, not in car sharing. This is the early and late majority, and their behavior change is required to bring this magic promise of self-driving cars to fruition.

Design work in support of cultural behavior change is, arguably, outside of one single person or company’s control. This is the gestalt of design, innovation, and technology movements. That gestalt (or zeitgeist) is emergent and constantly changing, and it takes an astute observer of all of the parts of the human condition to see that behavior from within, rather than in retrospect.

It is the aggregate of localized behavior change and habitualized behavior change that lead to this form of larger cultural evolution.

The autonomous vehicle hits on all levels of behavior change. At an immediate level, autonomous vehicles will have lots of interfaces, and that will require lots of localized behavior changes. We’ll need interfaces on our phones and in our cars that make it clear how we should be interacting with the technology. At a habitualized behavior level, we’ll need to become comfortable waking up in the morning and calling our car, or selecting our car of the day based on our mood. This isn’t harder than what we do now – but it’s different, and that difference will take consumer practice. And at a cultural behavioral level, we’ll all need to get our heads and hearts around changes to what has been ingrained in us as a fundamental: car ownership. Car ownership is built into the American psyche, as a car is an indicator of success, a point of pride, and for many, personified to the point of love.

Behavior change is about shifting things a person does at any given moment; about changing what a person does over time; and shifting what we all do – our cultural behavior. This model of behavior is useful for framing what we do in our product and service design jobs, how we do it, and our expectations for impact. The model gives us a way to zoom in and out of work, and to see how immediate and detailed design work ultimately supports larger, more fundamental changes to the world we live in.

Modernist can help you design for behavior change.

Get In Touch

Related Posts