Designing for Behavior Change: Applying Psychology and Behavioral Economics (2013)

Part IV. Designing the Interface and Implementing It

Chapter 10. Reviewing and Fleshing Out the Interface Designs

Once there is something real and tangible that the product team can look at—wireframes or mockups that convey the interface design—you can review them from a behavioral perspective and give feedback. Two different types of review should occur:

§ Looking for gaps in the overall user experience across the product

§ Looking for opportunities to apply specific tactics

Look for Big Gaps

The product design process is necessarily one of translation—converting concepts and intent into designs of something users can actually interact with. Along the way, the core aspects of the behavioral plan—the behavior change strategies used, the sequence of steps, the structuring of the environment and preparation of the individual—may change. Those changes can be highly beneficial from a behavioral perspective, if the designers were inspired to come up with new ways to meet the same ends. They can also be detrimental and miss key aspects of the application’s intended psychology.

Think of the interface design as a new behavioral plan, one that should be evaluated on its own merits and run through the design wringer we employed in Chapter 6 Chapter 7 Chapter 8. That way, the interface designs are not judged too literally against the original functional requirements if the UX designers came up with innovative behavioral solutions.

What You’re Looking For: The Create Action Funnel

At the end of the day, the product needs to help people change their behavior. It must put all of the prerequisites in place for people to act. We first talked about those prerequisites in Chapter 2, as the Create Action Funnel. Here’s a quick refresher on what needs to come together, at the same time, for a person to take action:

1. Cue. Something needs to start the person thinking about the action. It can be something already in the person’s daily life (whenever I see calla lilies, I think of hiking along the coast) or something the application does to catch attention (when I hear that particular buzz, I think about checking my text messages).

2. Reaction. The intuitive mind will automatically and rapidly react to the idea of taking the action. That reaction includes a basic sense about whether the action is interesting and pleasant, and it will also activate thoughts about other possible, related, ideas and actions. The product needs to get past this reaction without being rejected or having the person be distracted. I don’t answer my landline at home, for example, because I intuitively dislike telemarketers.

3. Evaluation. The conscious mind will evaluate the costs and benefits of the action, including the value the product provides and the frictions or other challenges in using it. That evaluation is relative, not absolute—the action must be sufficiently worthwhile and better, on net, than the other things that person is thinking about doing at that moment. I check my text messages, for example, because it’s easy, and the value I receive from them varies but is usually quite high.

4. Ability. The person actually has to be able to act right then. Potential barriers include not knowing what to do, not having what you need to act, or feeling that you’ll fail. These can be overcome, but that takes time. And when the barrier is resolved, the funnel starts anew—all of the other factors still need to be in place. For example, if I want to check my text messages, but can’t find my phone, I need to search for it. While searching, I might get distracted into a different line of thought.

5. Timing. There has to be a reason to act now, rather than doing something else that is more urgent. If I’m expecting a message from my wife telling me where to meet her in a few minutes, that’s urgent. But if the text messages can wait (and there’s something else I need to do, like a report that’s due), I’ll put it off even if I want to check the messages and have the ability to do so.

These five things are needed for the person to then execute the action; together they spell the acronym C-R-E-A-T-E. Habitual actions short-circuit the process; the evaluation and timing stages aren’t as important. But before the habit is fully formed, the user must still pass through the full Create Action Funnel.

As you review the interface design, ask whether each of these conditions is met for each small step the user needs to take (i.e., each page click and each form entry, along the way toward the final target action).

Fix any small, obvious problems you see—like the user not knowing what to do on the screen, or competing cues on a particular screen to do different things. There may also be larger problems, where there isn’t an obvious solution—like when there’s no apparent value to the user for taking a particular step on the way to the target action. In those cases, it’s time to return to the core design process: action-environment-user.

How to Fix the Big Gaps: Action-Environment-User

Let’s say you’ve identified large issues that need to be resolved. You can run through the design process from Chapter 6 Chapter 7 Chapter 8 to fix them. As a quick refresher, these tasks include the following:

§ Structuring the action to make it feasible for the user and to shift the burden of work from the user to the product where possible;

§ Constructing the environment, both the product itself and the individual’s broader decision-making context to motivate, cue, and give feedback to the user; and

§ Preparing the user to take the action through a combination of education, self-narrative, and building on existing positive associations.

These three tasks help you focus your attention on one part of the design process at a time so that you can best put the pieces in place for action. They provide guidance on “how” to go about the design process, after the Create Action Funnel provides “what” needs to get accomplished.

You can structure the process of generating solutions as a series of questions. Naturally, the particular questions you need to ask depend on the gaps in the application, but here are general ones to get you going:

Structure the action

§ How can these screens be tailored to better leverage the existing knowledge and experience of the users?

§ Do these screens present the Minimum Viable Action required of the user?

§ What parts can be dropped to simplify the process? What parts can be automated or intelligently defaulted?

§ Are they any repeated actions that can be turned into habits?

§ Is each action required clearly communicated to the user, and does it appear feasible up front from the perspective of someone with scant experience with the product?

Construct the environment

§ Does the user have a clear motivation to continue on each screen, and is she thinking about that motivation in that moment?

§ Are you asking the user to take a single clear action on each screen?

§ If the action takes longer than a few seconds, does the product provide clear feedback to the user of progress (or of the need to course-correct)?

§ What distractions are there on each screen that would pull the user away?

§ What distractions is the user likely to face given the time of day and location in which she will access this screen? How can you work around them?

Prepare the user

§ How does each user persona likely see himself, and does that self-concept support action on this screen?

§ Does this look familiar to what users already know, especially when it comes to taking action on the screen? Is it clear where and how to take action?

§ Does the user know what is needed, especially when action is required outside of the application? Are there clear, physically specific instructions?

Look for Tactical Opportunities

By now, the overall flow of the application should make sense. It’s time to fine-tune the content on each specific screen to maximize impact. Time and time again, behavioral economists and psychologists have demonstrated the significant effects that apparently minor changes in the presentation of a question can have on user responses; see the following sidebar for one of the many examples.

AN UNUSUAL DISEASE

One of the most famous examples of the power of small changes in wording is this study by [ref177]:

“Imagine that the US is preparing for the outbreak of an unusual Asian disease, which is expected to kill 600 people. Two alternative programs to combat the disease have been proposed. Assume that the exact scientific estimates of the consequences of the programs are as follows.

Program A: If Program A is adopted, 200 people will be saved.

Program B: If Program B is adopted, there is 1/3 probability that 600 people will be saved, and 2/3 probability that no people will be saved.

Which of the two programs would you favor?”

[And a second group was given this set of options:]

“Program C: If Program C is adopted, 400 people will die.

Program D: If Program D is adopted, there is 1/3 probability that nobody will die, and 2/3 probability that 600 people will die.

Which of the two programs would you favor?”

In the first group, people overwhelming selected Program A (72%). In the second group, people overwhelming selected Program D (78%).^[122^] All of the programs are exactly the same in terms of the expected number of people that will live and die. One of the things they do differently is present the program as a loss or a gain.

Losses are a tremendous motivator for behavior, as described later in “Leverage loss aversion.” If you can frame the target action as avoiding loss, without becoming too negative overall, you can greatly increase the chance of action.^[123^]

The best strategy to change behavior can succeed or fail because of the specific tactics the application uses (i.e., how, exactly, the application interacts with the user). It’s essential to take into account how our minds react to content on the screen.

At first, the range of possible tactics can seem overwhelming. In Chapter 1, I mentioned that there are hundreds of different mechanisms that researchers have identified that affect how we make decisions; most can affect how the specific content is put in applications. Many of the books that already exist on psychology and the design process focus on these tactics—from Stephen Anderson’s Seductive Interaction Design (New Riders, 2011) to Dan Lockton’s website Design with Intent Toolkit (http://www.danlockton.com/dwi), to Susan Weinschenk’s 100 Things Designers Need to Know (New Riders, 2011).

We need a way to organize the impact that these mechanisms have, to categorize them and determine if that’s what’s needed on a given page.^[124^][125^] The Create Action Funnel from Chapter 2 provides a useful structure for that purpose. You can think of each of the individual page-level tactics as affecting one or more stages of the Create Action Funnel—making it more (or less) likely that the person will take action on the given page.

So, here’s a scenario. Your users have just loaded a particular screen on your website, web app, or mobile app. This screen represents one of a sequence of steps that lead up to the target action that the user wants to take (or there’s only one step, and that’s the target action). How can you speed the person through that particular page? Table 10-1 presents two dozen tactics you can use, organized by the part of the Create Action Funnel that they affect most strongly.

Table 10-1. Tactics to support action on a particular page

The following sections describe each of these cognitive mechanisms, and how you can deploy them to the user’s advantage. While I’ve devoted a separate section (this one) to how you can apply behavioral tactics to an existing interface design, this should ideally occur simultaneously with the main interface design process.^[126^]

Tactics for Cueing

Many of the tactics listed here have been mentioned earlier in the book, under the discussion of behavior change strategies or the conceptual plan. In those cases, the text focuses on how that tactic can be employed, in particular, at the interface design level. The goal of this section is to provide a quick reference to each of the major tactics you can use to improve your interface designs, all in one place.

Tell the User What the Action is (and Ask for It)

See also: Chapter 2, Chapter 7

Cueing users to act on a given page starts with something that should be obvious, but unfortunately isn’t: making sure they know what action is desired. I know, you can’t imagine that anyone would make such an obvious mistake. But we all do, all the time. Do you include a link to your website at the bottom of your emails? If so, do you actually ask people to look at your site, or do you hope it’s obvious? Do you post your Twitter handle on your messages or blog posts, hoping people will follow you? Readers could figure out the action we intend (view website, follow on Twitter), sure. But the more mental leaps that are required between what we see (Twitter name) and the action, the less likely is it that the action will cross our minds before we’re distracted by something else.

[ref40] ran a set of experiments on how he presented his Twitter handle to readers of his blog. He started with a simple informative statement: “I’m on Twitter,” in which “Twitter” was a link to his page, and 4.7% of readers clicked. Then, he did the obvious—which apparently isn’t so obvious to the rest of us—he told people what the action was: “Follow me on Twitter.” Boom—7.31% of users clicked. And, even clearer: “You should follow me on Twitter here”—12.81% of users clicked. There are multiple effects at work in the last statement (a personal request, specificity, etc.), but the effect of requesting the action is undeniable. One lesson is simple: directly, and unabashedly, ask people to take action.

Make It Clear Where to Act

We scan, we don’t read. Don’t expect users to read lots of text on your page. The two-second rule is a good test—if you don’t get the gist in a two-second glance at the page, you risk losing the reader’s attention. Krug’s Don’t Make Me Think (New Riders, 2006) gives a great overview and practical examples, and Johnson’s Designing with the Mind in Mind (Morgan Kaufmann, 2010) talks about the visual perception system and related psychology.

Some of the key things that we quickly recognize are the ways in which we can interact with a page (affordances per [ref141])—what looks like it is clickable, doable, or can otherwise get you off this page and quickly on to the next one. The lesson is simple: make buttons look like buttons, and make anywhere else people are expected to take action clearly a place where they can take action.

Clear the Page of Distractions

See also: Chapter 2, Chapter 7

Since we’re scanning, and trying to save work, we’re all too likely just to click on the first thing that looks clickable. Make a single clear call to action, if your goal is to get the person to keep moving through the page. Remove extra links and buttons, or place them in a distinctly lower level in the page’s hierarchy.

Bonus Tactic: Blinking Text

Blinking text is a really great cue. It never fails to catch our attention! And it’s also darned annoying because it catches our attention and won’t let go. Don’t do it. Seriously.

Tactics Related to the Intuitive Reaction

Make the Site Professional and Beautiful

We rarely consider unprofessional-looking websites to be credible ([ref72]). If you’re trying to help someone take an action, don’t make them have an intuitive reaction of distrust. That’s unnecessary friction on the action. Like it or not, we assume that scammers make bad websites and apps. We even find it easier to use clean, well-designed products (see [ref6] for examples). So, even if your app is created to help people, it has to look good.

Deploy Social Proof

If we see that other people are taking an action, we’re more likely to feel that the action is valuable and worthwhile. It’s a quick gut check—if everybody does this, it must be OK, right? This is one of the major ways that our minds save work and quickly make decisions in uncertain situations.

Using social proof is a key tactic in sales and persuasion, with a long research tradition behind it ([ref33]), and can be valuable for page-level behaviors as well. You can convey the fact that other people are taking the same action by using people’s faces, short testimonials, or even a brief statement like “90% of people finish this step” (this is similar to a peer comparison, explained later). For more information on this topic, refer back to Chapter 1.

Display Strong Authority On the Subject

People are more likely to trust those who they see as an authority on the subject. If you’re telling your users that they have to do action X in order to build up to their goal of Y (and it’s true), then speak with authority. Don’t write wishy-washy text. Make sure your credentials can be seen without beating your users over the head with them.

There are great studies on how people wearing suits, or with professional titles, are simply assumed to be more credible and trustworthy. The use of (perceived) authority is also a favorite tactic in sales and persuasion. See Cialdini’s discussion of the underlying research (2001).

Be Authentic and Personal

People pay more attention to personal appeals to act than to impersonal ones. If you receive a letter with a hand-written envelope, how likely are you to open it? How about one with a standard machine-printed address? The reasons are manifold, but we are more likely to ignore machine-generated, impersonal appeals than tailored, personal ones ([ref76]; [ref139]). We have an almost automatic response of “this is spam” for any email or letter from impersonal sources.

Here’s a great example of using personalization and authenticity to cut through the noise and get people’s attention. In Oregon, there’s a lottery for free healthcare for people that can’t afford it. But some of the people who sign up for and end up winning the lottery don’t open the letters notifying them that they’ve won. And so they miss their chance at free healthcare.

Ideas42, the leading behavioral economics consultancy in the United States, devised a simple outreach campaign to the winners of the Oregon healthcare lottery. They notify people that they’ve won with a postcard featuring the smiling faces of the people at Providence Health in Oregon who will help them sign up for their healthcare. The recipient’s name and address is handwritten on the postcard. Figure 10-1 shows a sample.

Figure 10-1. A postcard developed by ideas42 to help winners of the Oregon healthcare lottery get past their automatic rejection of form letters, and read enough to see that they’ve won free healthcare

Remember, we’ve been conditioned to reject impersonal and computer generated appeals. Most of us have an intuitive reaction against them. To avoid that intuitive reaction, our products need to do something different, something that’s good practice anyway: be authentic and personal.

Tactics Related to the Conscious Evaluation

Prime User-Relevant Associations

We each have multiple possible mindsets with which we interpret and respond to the world. Those mindsets are selectively activated, based on our (very) recent experiences.

If you’re asking people to commit to running once a week, get them thinking about previous times they’ve run, first (as long as that experience was positive)! When you ask them to commit to running, the benefits of running will be clearer and more salient in their minds. For more information on this topic, refer back to Chapter 1.

Leverage Loss Aversion

People respond much more strongly to losses than to gains—they are “averse” to losses. In fact, in many scenarios, people will be willing to forfeit twice as much money to keep an item that they already have (and have no other personal attachment to), than they are willing to pay to purchase an otherwise identical item. There’s a detailed literature on special cases of loss aversion, but that general rule holds true in many cases: losses are roughly twice as motivating as gains ([ref100]).

Loss aversion is a very powerful tool to help people change their behavior. By selectively framing the presentation of a desired action as avoiding loss rather than gaining benefits, the application can trigger a strong gut reaction to act. For example, it’s can be much more persuasive to tell a guy that he’ll lose sexual potency unless he gets in shape ([ref109]), rather than telling him he’ll gain a more attractive stomach.

When leveraging loss aversion, though, remember that your users can just stop using your product to avoid loss and negative emotions that come with it. The product must be seen as worthwhile and enjoyable overall—loss aversion should only be used on the margins.

Use Peer Comparisons

Being told about, and being compared to, the actions of our peers can be immensely powerful. Our behavior frequently conforms to what we believe our peers do (i.e., “descriptive norms”), compounded by the usually false belief that our peers are watching our behavior and judging it (“spotlight effect”). This effect has been shown in everything from energy usage ([ref35]) to voting ([ref78]).

For behavioral products, the implications of peer comparisons are tremendous. Social norms are an incredibly powerful part of one’s micro-environment and can encourage (or discourage) action. The same is true within the context of each individual screen that the user interacts with.

To use this technique, compare users’ performance to a reference group that they care about (their friends, colleagues at work in a similar job position), and try to ensure that the reference group you choose is doing better than the user. Peer comparisons encourage people to move toward the norm (the average for the reference group). So, if you tell them they are already doing better than most people, they’ll just relax and not work so hard. But that negative effect can be counteracted with an explicit social approval (“great job!”) for exceeding the norm ([ref158]). For more information on this topic, refer back to Chapter 1.

Use Competition

We all have a natural competitive side—though it’s much stronger in some people than others. Usually, you’d build a competition into the overall product (as described in Chapter 9, as one of the possible “design patterns for behavior change”), but it can be deployed at a page-level, too. For example, imagine a page that has people match Spanish words to their English meanings, to help the users learn Spanish. The page could include a counter of how many correct answers the individual has versus others on the page at the same time.

Avoid Cognitive Overhead

Make it straightforward and clear what the user needs to do: each time the user has to make a logical leap from, “Oh, if I do this, then this will probably happen, but I’m not sure,” that’s costly. It takes time and energy away from the task at hand.

[ref116] gives a great example of a product that is physically easy to use, but still is costly to the user because of cognitive overhead. Here’s his hypothetical user thinking through a QR Code: “So it’s a barcode? No? It’s a website? OK. But I open websites with my web browser, not my camera. So I take a picture of it? No, I take a picture of it with an app? Which app?” Forcing your users to think about what to do should be reserved for cases where their input is important and will shape their outcomes; don’t force your users to expend energy because the product is confusing.

Avoid Choice Overload

A growing body of work demonstrates the difficulties individuals face when confronted with too many choices. Despite the common wisdom that more choices are better, two problems arise. First, people may refuse to make any decision at all. Second, people may regret the choices they made in an impossible search for the “optimal” choice ([ref94]; [ref159]).

For example, an often-cited study by [ref96] placed two different displays of jam in a grocery store: one with 24 jams, and one with 6. The 24-jam display attracted 60% of customers, but only 3% of those shoppers ended up buying any of them. The 6-jam display attracted 40% of customers, but 30% of them bought one. Subsequent studies have also shown that satisfaction with one’s choice, whatever it is, decreases with the number of options one had to choose among.

There is an obvious implication here when constructing individual pages in an app—avoid situations in which the user has to choose among a large number of options (if you want the user to make a choice and be happy with it). There is also a less obvious lesson: be wary of users (and fellow employees) who say they would really like more options. The person is probably telling the truth, at least from the perspective of that person’s conscious deliberative self, but that doesn’t mean providing more options is the right thing to do.

Avoid Direct Payments

You can always pay people to click on your button. But I don’t recommend it. If you need to pay people to do something that’s supposed to be a voluntary behavior change, you’re probably not connecting that small action with the reason they want to change their behavior in the first place.

There is extensive evidence that financial incentives induce individuals to undertake behaviors that they would not undertake ([ref98]). People are motivated by money. No great surprise, right? However, when a person is already inclined to take the action, financial incentives can backfire by decreasing preexisting internal (intrinsic) motivations ([ref81]); the individual is more likely to stop the behavior after the incentive is removed. Similarly, other social, motivations are crowded out (see [ref8] for examples) when we start thinking about our behavior in terms of being paid to act. Direct payments are less likely to cause problems with once-off behaviors, like signing up for the gym. But they can undermine long-term intrinsic motivation—like actually going to the gym over time! For more information on this topic, refer back to Chapter 7.

NO MAGIC WANDS

Throughout this book, I provide the tools you need to find the behavioral processes and product features that work in your particular context and verify their impact with your specific set of users.

What I can’t do is give you the secret behavioral tricks that will always change user behavior in predictable ways. That’s because such magical formulas simply don’t exist (run away from people who tell you they do!). All behavior change interventions interact with an individual’s desires, prior experiences, personality, and knowledge to produce their unique impact for that person. There is just too much variation across people for any approach to always work.

Most of the approaches and lessons that I talk about here have been tested either in a researcher’s laboratory or in a specific product setting. In most cases, I’ve also observed these techniques in practice in my own work, or through the dozens of companies I’ve interviewed and learned from. Unfortunately, though, there are very few studies out there that apply and rigorously test theories of behavior change in ways that can be generalized to lots of other products. That’s something we strive for at HelloWallet—but even then, it’s very difficult to make the case that what works for us, in helping people take control of their finances, is going to work the same way for someone else’s dieting software.

We are all still in the early stages of learning how to use products to help people change their behavior. So, in Chapter 14, I provide some guidelines on how to test specific interventions in your product and move the field forward at the same time. I encourage you to contribute your findings to the broader community so we can all learn and develop our skills together.

Tactics Related to the User’s Ability to Act

Elicit Implementation Intentions

As you may recall, implementation intentions are specific plans that people make on how to act in the future ([ref84]). They are a form of behavioral automation, telling the mind to do X whenever Y happens.^[127^] The person does the work of thinking through what needs to be done now, and then when the action is actually needed, there’s no need to think, and no logistical barrier to action—the person just executes the action. Implementation intentions should include the event that triggers action, the context for that action, and the physical things the person should do. For example: “On Friday at work, if my supervisor yells at me about the project, I’ll leave the room and take a short break rather than yelling back.”

You can encourage the user to create a future action plan (implementation intentions) wherever the user is committing to take some future action, especially when that action is outside of the application. Making a specific, concrete plan of attack can help the person follow through with the action, even when the product isn’t there as a reminder.

For behavioral products, deploying implementation intentions can mean adding text boxes where users describe how they’ll take the action. The key is to make people think consciously about the concrete actions, and, if possible, visualize undertaking those actions. For more information on this topic, refer back to Chapter 7.

Default Everything

The power of defaults was touched upon earlier, when planning the overall behavioral strategy of the application. But it is also important to keep in mind for individual input fields within an application. In short, assume that many users will stick with whatever default value you give them. This occurs because people are in a hurry and don’t fully read the questions posed to them, because they are unsure of what the question means, or because they simply do not have a strongly held preference.

This means that default values can be immensely useful (a) where the default response can move the individual closer to action, (b) where power users can fine-tune their responses, and (c) where everyone else can breeze past the defaulted values. However, default values should be used only where nonresponse is acceptable; it shouldn’t be used where essential information is gathered. And, since users will make up fake answers (or simply disengage) when forced to answer questions that they can’t really answer, it is better to altogether remove questions that users don’t have answers to and can’t be defaulted. When default values are provided, the answers should be interpreted as one part truth, and one part nonresponse.

For example, let’s say your application asks users if they have kids. If there’s special advice that’s only relevant for people with kids, then default the answer to “no kids.” Let those users who do have kids, and are paying enough attention, indicate it to receive the special content. For more information on this topic, refer back to Chapter 3.

Lessen the Burden of Action and Information (Cheat!)

Part and parcel with defaulting is removing the need for users to do extra work. That’s a high-level behavior change strategy, and it can and should be used as a tactic for particular interactions as well. If you don’t need to ask a question of the user, don’t. If you can save the user from scrolling down the page, excellent. That’s just another small, but frictionful activity that the user needs to take on their path to action. Removing these frictions mean decreasing the cost of action, all else constant.

That isn’t to say that users can’t do work. There may be really important information below the fold, and the user really does need to read or act on it. However, if there is a choice between accomplishing the same task with or without additional form fields and user work, choose the route with less work. For more information on this topic, refer back to Chapter 3.

Deploy Peer Comparisons

If we think a task isn’t achievable, we have better things to do with our time. On your pages, make sure not only that the person can do what’s needed, but that they know they can do it, too. One way to accomplish that is through the peer comparisons described earlier: show the user that other people are successfully taking the action. For more information on this topic, refer back to “Use peer comparisons.”

Tactics Related to When the Timing Is Right for Action

Ideally, the action on the page is inherently time-sensitive: people need to take the action immediately because of some existing, external rationale—like taxes on April 15. However, when that’s not possible, there are a few other tactics that you can use.

Frame Text to Avoid Temporal Myopia

We’re wired to value the present far more than the future—that’s our “temporal myopia.” We talked about that earlier, when we looked at ways to motivate the user with immediate rather than future rewards. Well, what if you’re stuck, and the basic structure of the application and its core motivation is already fixed? You can still avoid the curse of temporal myopia by crafting the descriptions you provide to the user.

When designing for behavior change, this means being very careful about the framing of time. Look for ways to frame benefits in terms of immediate or near-term gains; in the weight-loss example, reference how the dieter will feel and look better almost immediately. The opposite is true forpain and effort: effort that occurs sometime in the future is much easier to commit to than effort right now. So, if the pain and effort needs to be discussed at all, put it in the future as much as possible.^[128^] [ref17] do this beautifully with their Save More Tomorrow plan—people commit now to saving (i.e., pain) at a future date. For more information on this topic, refer back to Chapter 7.

Remind of Prior Commitment to Act

We don’t like to be inconsistent with our past behavior. It’s very uncomfortable, and we have a tendency to either act according to our prior beliefs, or change our beliefs so that they are in line with our actions ([ref62]). One way to achieve this is to have users impose urgency on themselves—promise to take the action at a specific time, then come back to them and remind them at that point. In addition to their other reasons to act, that will spur them to follow through, to avoid feeling inconsistent. For more information on this topic, refer back to Chapter 7.

Make Commitments to Friends

Another way to create urgency to act is to make specific promises to do so to one’s friends. Social accountability is a powerful force—we don’t want to let our friends down, or lose esteem in their eyes. The sidebar illustrates that point with a personal story from my friend Justin Thorp.

Our friends have a wide range of effects on our behavior, as we’ve talked about previously under the power of social proof and descriptive norms. But telling our friends what we’re doing has a particular power to push us to act when we say we will. It’s not just the action by which they judge us, but whether we kept our word overall—and that includes timing.

OUR FRIENDS HOLD US ACCOUNTABLE

I’ve always been a big guy. Back in the fall of 2009, I was clocking in around 280 and was getting fed up with being winded when I ran up a few flights of stairs. I knew it was time to do something different.

So I decided to get into running. That was a daunting task for me. I could barely run down the block. Being a nerd, my first thought was, what’s better than exercising? It’s exercising with technology. So I perused the app store and got RunKeeper—an app that used my phone’s GPS to track how far, how fast, and where I ran. I was instantly hooked. It allowed me to see my progress throughout my running journey.

After a while, I noticed little Facebook and Twitter share buttons at the bottom of the RunKeeper report. With the press of the button, I could share my runs with my friends. I was like, “What the hell?” and hit the button not really thinking much of it.

A few days passed, and all of a sudden my friends and coworkers started noticing my runs. They were commenting on my Facebook posts. They were cheering me on via social media. When I wouldn’t run, my boss would ask, “Justin, why didn’t you go running today?”

When I got up in the morning and didn’t want to go running, I’d hear the voices of my friends and supporters in my head. I didn’t want to let them down. They believed in me and believed that I could do it.

And I did. I lost 50 lbs. I ran the Cherry Blossom 10 Mile Race. I gained a ton of confidence. And I still run regularly. It’s become a great way for me to get exercise and clear my head.

—Justin @thorpus

Of course, it all depends on who we look to for support and accountability. If we turn to people who really don’t care about us, or who don’t value the activity we’re trying to undertake, then their disinterest can sap us of motivation. Products can mitigate this by explicitly asking people to identify friends and colleagues who will support them, or by matching up the person with other users who are seeking to change the same behavior or have experience providing support (i.e., products can construct a local network of peers who will push us to succeed). Lift.do does something akin to that, as does Goal Sponsors, for a fee.^[129^]

Make a Reward Scarce

You can make a reward for the action scarce (“the names of the first 100 people losing 100 pounds will be featured on our website”), or artificially time-sensitive (“act in the next five minutes, and you’ll get another 10 points”). This is another favorite sales and marketing tactic ([ref33];[ref2]). It’s best for once-off actions, and not repeated behavior. If you try to repeatedly push for a behavior with scarcity, people will stop believing you. Also, you run the risk of desensitizing the person to normal scenarios that aren’t artificially scarce or time-sensitive.

Update the Interface Designs

After the interface designs have been evaluated for large gaps, and opportunities to leverage specific cognitive mechanisms have been found, you should make recommendations to the product manager for changes to the designs. These changes would then be considered in terms of their behavioral, engineering, and aesthetic impact. The end result is an updated interface design.

On a Napkin

Here’s what you’ll need to do

§ Review the interface designs for the five prerequisites: cue, reaction, evaluation, ability, and timing.

§ Apply specific tactics in each of these areas, leveraging the literature in behavioral science.

§ If there are large gaps, you can repeat the exercises from Chapter 6 Chapter 7 Chapter 8, focusing on the action, the environment, and the user in turn.

How you’ll know there’s trouble

§ Endless debates over wording and design details. Such questions should be tested in practice.

Deliverables

§ Revised wireframes or mockups