a toolkit for designing for behavior change - Action Design - HelloWallet

A TOOLKIT FOR DESIGNING FOR BEHAVIOR CHANGE

TIPS FOR APPLYING TECHNIQUES FROM THE BOOK, DESIGNING FOR BEHAVIOR CHANGE

A Toolkit for Designing for Behavior Change

A Toolkit for Designing for Behavior Change Contents Introduction ....................................................................................................................................... 2 Overview ............................................................................................................................................ 3 Step 1: Understand............................................................................................................................ 4 Step 2: Discover ................................................................................................................................ 5 Step 3a: Conceptual Design............................................................................................................. 6 Step 3b: Interface Design ................................................................................................................ 7 Step 4: Refine ..................................................................................................................................... 8 Teaching the Create Action Funnel ................................................................................................ 9 Quickly Evaluating a Proposed Feature with the Create Action Funnel .................................. 9 Conducting the Discovery Phase for a New Product ............................................................... 10 Conducting the Discovery Phase on an Existing Product ........................................................ 12 Developing the Conceptual Design ............................................................................................. 13 Illustrations of the Process............................................................................................................ 15

Page 1

A Toolkit for Designing for Behavior Change

I NTRODUCTION

This toolkit is designed to help you design products that enable your users to change behaviors in their daily lives – from exercising more, to saving for the future. It’s meant to accompany the book Designing for Behavior Change, which will be published by O’Reilly Media in November 2013. The toolkit provides a brief summary of the book’s key lessons, and gives additional tips on how to apply the book to your own projects. Both the book and this toolkit are based on our experiences at HelloWallet. Over the past few years we’ve successfully built and experimentally tested products that help people take control of their finances. But, we had numerous mistakes and outright failures along the way – as we searched for effective ways to apply the behavioral science literature to software products. We also had the unique opportunity to share notes about behavior change with the folks at Action Design DC, 1776 DC, and 500 startups where we’ve field tested these ideas; we are indebted to them for them help. In the end, we’ve codified these lessons in a step-by-step process for discovering, designing, implementing, and iteratively improving products that help their users take action. That full, highly detailed process, is given in Designing for Behavior Change. In this toolkit, we’ve tried to create the shortest, simplest presentation of main themes from the book, but necessarily had to drop important details. If you’d like to learn more, please check out the full book on Amazon or Oreilly.com.

NOTE -- This toolkit is very much still a draft, and we welcome your feedback. You can reach out to Steve Wendel via “sawendel” on Twitter, LinkedIn or Email (sawendel at hellowallet.com) with any comments or suggestions.

Page 2

A Toolkit for Designing for Behavior Change

O NE | D ESIGNING

FOR

B EHAVIOR C HANGE

AT A

G LANCE

Over view Designing for Behavior Change is a four-phase process: 1. Understand how the mind decides to act and what that means for behavior change 2. Discover the right behaviors to change, given your goals and your users’ goals 3. Design the product itself around that behavior 4. Refine the product’s impact based on careful measurement and analysis These four phases layer on-top of, and don’t replace, your existing product development process. You can use an Agile Process, Lean, or even a sequential development (Waterfall) process. At some point in each development methodologies, the tasks of understanding, discovering, designing, and refining around a particular behavior slot in. Figure 1 shows an example of that process, using an Agile development process. The dots show the outputs of each stage. In this Toolkit, you’ll learn how each of these outputs work.

The Four-Step Process

When you already have a product in the market, and want to improve it using behavioral techniques, the process is similar. After an initial understanding of the decision making process, and clarification of the product’s outcomes, you quickly jump to the refine phase: assessing the current impact of the product. Then, as needed, you do additional discovery, design, and further rounds of refinement.

Page 3

A Toolkit for Designing for Behavior Change

O NE | D ESIGNING

FOR

B EHAVIOR C HANGE

AT A

G LANCE

Step 1: Understand The first phase entails understanding how the mind makes decisions. In order for a person to take action, five preconditions need to be in place. Here they are, using the example of a running app like RunKeeper: •

Cue: Something needs to cue the person to think about acting. Why would your users think about running? Maybe a push notification, an SMS invite from a friend, or an ad on TV.

•

Reaction: The mind automatically reacts intuitively and emotionally. What do your users think about running? For some, it’s great; for others it’s new and strange, or embarrassing to be out of shape.

•

Evaluation. With conscious awareness, the mind does a quick cost-benefit analysis. How hard will the action be to take, what’s the action’s value for the user, what are other alternatives, etc.? For some users, running is a net negative (maybe it aggravates a knee condition) and for others, positive.

•

Ability. The person must actually be able to act and know it. The person must know logistically what to do, have the resources to do it, and not be dissuaded by an assumption of failure. Some users may not have running shoes, for others it’s raining outside: they can’t run even though they want to.

•

Time pressure. The person needs to have a reason to act now, rather than doing something else that is more urgent. The user may want to run, but is busy doing something else. If all of these are in place, then a person can execute the action. Together, they form the C-r-e-a-t-e Action Funnel. At each point, people leak away: either intentionally deciding not to take the action, or becoming distracted by other things. Designing for Behavior Change helps people pass through all five stages, from inaction to action. If the user faces obstacles, often they can be resolved; but, that process takes time and leaves the person open to distraction along the way.

The Create Action Funnel

There are three basic strategies to pass this funnel: helping users make a conscious choice to act, triggering an ingrained habit, or restructuring the action to “cheat” so the product performs the action on the users’ behalf, given consent. Each of these start with a conscious process, and so we focus on that here (all three are covered in the full book). Page 4

A Toolkit for Designing for Behavior Change

O NE | D ESIGNING

FOR

B EHAVIOR C HANGE

AT A

G LANCE

Step 2: Discover Designing for Behavior Change builds upon a clear understanding of the target outcome, action, and actor: •

The outcome is what will be different in the real world when the product is successful;

•

The actor is who will cause that change: the actor is usually the user of your product; and

•

The action is how the actor will do it – the behavior the actor will undertake.

Let’s say your company has identified a market opportunity: people want to be healthier, and aren’t satisfied with current products. Here’s how to process would go: 1. Define the outcome. Make sure it’s observable. Avoid states of mind like “knowing how to eat healthy” that fall short of the outcome you really care about. E.g. Weight loss. 2. Identify the Actor. Who is your target market? The clinically obese? Active 20-25 year olds? Be specific as possible. E.g. Sedentary office workers in New York City. 3. Select the Action. Brainstorm at least 5 very different ideas about what action your actors can take that will meet the outcome (for example: healthier eating, a diet pill, gastric bypass, exercise). Then, evaluate according to four criteria: a. Impact: How well will the action actually achieving the target outcome? b. Ease: How difficult is it for the user to take that action? c. Cost: How costly will it be for you to build a product supporting that action? d. Fit: does the action make sense for the company’s larger goals and culture? For example: a diet pill may be effective, but is very costly to develop and doesn’t fit the culture of the company. Let’s say you decide on the action of running.

The goal of the discovery phase is two-fold: (a) to provide guidance what to design and build and (b) to fail fast. You can find problems early in the development process: problems where the team disagrees on the product’s actual purpose, who the target audience is, how the product will be judged for success or failure.

Page 5

A Toolkit for Designing for Behavior Change

O NE | D ESIGNING

FOR

B EHAVIOR C HANGE

AT A

G LANCE

Step 3a: Conceptual Design In the design phase, we deploy a considerable body of behavioral research, at both a conceptual level (what functionality the product should have), and at the interface level (what the product should look like). At each level, we try to change the entire decision-making context to support action; the product itself is just one part. That context encompasses the user, the surrounding decision making environment, including the product itself, and action the user is taking. Here’s how we shape each one, for the conceptual design: •

•

•

Structure the Action. Break the target action down into discrete steps. Refine the steps using a combination of automation, simplification, and tailoring to the users’ exiting expertise and knowledge. For example, develop a series of runs of increasing difficulty for the user to train on. Focus on the Minimum Viable Action: the smallest user action that can test the product’s viability. Design the Environment. Design the product itself and broader decision environment to support action by ensuring the motivation and cue for action is clear, the user gets clear feedback on progress, and competing actions are sidelined. For example, a tracker on the user’s wrist provides clear feedback and reminds the user to go running. Prepare the User. Use a combination of narration (change how users see themselves), association (build upon other things they’ve done before), and education (show them logistically what’s required to complete the action). For example, have the user think about themselves not just as someone who runs, but as a “runner” who’s increasingly fit and healthy.

Use this three step process, focusing on the behavior, then the environment, then the person, to develop a behavioral plan – a narrative that describes what the user might do, to progress from not taking the action, to taking it. That plan can be articulated in a customer experience map, journey map, textual narrative, or a simple list of bullet points with annotations. Then, for each step in the journey, use the Create Action Funnel as a checklist, to ensure you have all five pieces in place for action (cue-reaction-evaluation…). The three design elements (action, environment, and user) are how you influence behavior; the Create Action Funnel tells you what needs to happen here.

Page 6

A Toolkit for Designing for Behavior Change

O NE | D ESIGNING

FOR

B EHAVIOR C HANGE

AT A

G LANCE

Step 3b: Interface Design After you develop the conceptual design, translate the concepts into the raw material of your team’s product development methodology. In an Agile environment, that’s User Stories. In a sequential development environment that’s functional requirements and specifications. The team will take the stories or requirements and development wireframes or clickable prototypes. But, avoid taking the behavioral plan as a simple blueprint for the actual UI: nobody likes a to-do list! The UX team must be free to innovate within the functional constraints given by the behavioral plan’s user stories. With wireframes or prototypes in hand, get of the building and get in front of users. Identify where problems lie, using the Create Action Funnel as a guide (e.g.: do users have a strong, negative emotional reaction?). When you identify the problem, you can deploy tactics from the behavioral literature to encourage action, and resolve those particular challenges. Here are the most powerful, general purpose tactics: Component:

To Do This:

Try This:

Cue

Cue Action Increase Power of Cue Increase Power of Cue Increase Trust Increase Interest & Trust Increase Interest & Trust Bypass Automatic Rejection Increase Motivation Increase Motivation

Tell the User What the Action Is Make It Clear Where to Act Clear the Page of Distractions Make Site Beautiful and Professional Deploy Social Proof Display Strong Authority on the Subject Be Authentic and Personal Prime User-Relevant Associations Leverage Loss Aversion

Increase Motivation Increase Motivation Decrease Cost of Action Decrease Cost of Action Increase Motivation

Use Peer Comparisons Use Competition Avoid Cognitive Overhead Avoid Choice Overload Avoid Direct Payments

Increase Logistical Ability Decrease Constraints

Elicit Implementation Intentions Default Everything

Decrease Constraints Increase Sense of Feasibility (Self-Efficacy)Constraints Increase Urgency Increase Urgency

Lessen Burden of Action and Information (Cheat) Deploy (Positive) Peer Comparisons Frame text to avoid temporal myopia Remind of prior commitment to act

Increase Urgency Increase Urgency

Make it scarce Make it time-sensitive

Reaction

Evaluation

Ability

Timing

Based on the (revised) interface designs, you then build the product itself. Page 7

A Toolkit for Designing for Behavior Change

O NE | D ESIGNING

FOR

B EHAVIOR C HANGE

AT A

G LANCE

Step 4: Refine Whenever we make products that interact with the vagaries of human behavior, we’re going to get some things wrong. The design process makes the initial product less wrong. But, there will always be significant room for improvement. Improving the product proceeds like this: 1. Assess the impact. Start with clear metrics of what you’re trying to accomplish – the target outcome and action from the Discover step. Before you measure anything, define what “success” and “failure” look like – to avoid wrangling and spin-doctoring after the fact. Then, it’s time to measure: a. A/B Tests. If possible, run an A/B test or another type of controlled experiment to measure impact. That’s the gold standard, and there are numerous tools to help companies run them with online products. Statistical knowledge is not required (though it can help). b. Other Options. If experiments aren’t feasible, you’ll need to measure overall impact, then formally or informally control for other things that might influence user behavior. Statistical techniques such as matching and panel analysis can mitigate these issues, but require statistical expertise. c. Build a data bridge. If the outcome isn’t directly measurable within the product, develop a data bridge: a statistical relationship between things you can measure in the app, and the real-world outcome. (E.g.: when people say they’ll go running, 34% of the time, they actually do). It can come from existing research, or a small pilot study conducted by the team or third party academics. 2. Develop Insights and Ideas to improve the product. Watch your users using the product. Look for page-by-page drop-off in your data. Segment the user population and see who the product is helping, and who it isn’t. All of these are techniques to find where problems occur. Sometimes, why they occur is obvious (the page is broken); at other times, you can do additional user testing or use the Create Action Funnel to think through the types of problem users face. 3. Change and Measure the impact of each change on your users. a. Prioritize proposed changes to the product designed to improve behavioral impact alongside other proposed changes meant to support business goals, engineering goals, etc. b. Test each major change. Regardless of whether everyone on the team thinks it’s a good idea, check the impact of that change on behavior. Human behavior is just too complex to accurately forecast, and a culture of testing can help check everyone’s assumptions.

Page 8

A Toolkit for Designing for Behavior Change

TWO| E XERCISES Below are a set of VERY ROUGH exercises that you can use for your team to apply this method; feedback welcome! We’ll do a well-design set of exercise sheets next. For each of the exercises below, divide into groups of up to 8 people. Each group should have pens and a shared writing space (whiteboard, butcher paper, etc.).

Teaching the Create Action Funnel 1. List out two or three products that team members have worked on, such as email campaigns, mobile apps, or websites. 2. For each product, write down: a. The target action that the product sought to encourage. The team doesn’t need to be too specific about the action at this point (we’ll get to that later). b. The Cue required for action. What would trigger the person to take the target action, including logging into the application, if needed? c. The emotional and intuitive Reaction you’d expect from users when they start thinking about the target action. d. The conscious Evaluation of costs and benefits you’d expect from users. e. The Ability of user to take the action. Can they actually do it, while they are interacting with the product? Or do they have to remember to do it later? f. The Timing or Urgency for users to take the action. Do users have a specific reason to take the action immediately, rather than waiting until some later date? 3. For each product, ask: where did the product falter? How could it have been improved?

Quickly Evaluating a Proposed Feature with the Create Action Funnel When the team already has a product or feature in mind, you can quickly stress-test the feature using the Create Action Funnel. Use a quick checklist, like this: Cognitive Phase

Proposed Feature

Alternative Feature (Optional)

Cue to think about taking action Emotional Reaction Conscious Evaluation of costs and benefits Ability to act (resources, logistics, self-efficacy) Timing and urgency to act Add notes in the margins for more information, as needed. Page 9

A Toolkit for Designing for Behavior Change

TWO| E XERCISES Conducting the Discover y Phase for a New Product The goal of this exercise is to determine the aims of the product -- outcome, action, and actor a new product should target.

Step 1: Outline the project brief 1. Jot down core facts about the market opportunity or need for the product. For example, “our audience wants to exercise more, but they dislike existing products because…”. “HR wants to enroll members in a new wellness program and needs to develop a compelling communication”. 2. Jot down external requirements placed on the product. I.e., “for business reasons it needs to serve people nearing retirement age”. “It needs to work on iPhones”. Only write down issues that are true requirements, and not assumptions about the audience or how to best reach them. 3. Write down what you hope to achieve with the product. What’s the real-world measureable outcome? a. If you naturally focus on outcomes related to your company (“doubling our market share by appealing to 25-35 year olds”), that’s fine. We’’ll call that your company objective. After recording that, write down why your users would care. What real-world change would they see and value? That’s the target outcome. b. Define a rough metric for how you’ll know the product is successful or not at achieving the target outcome. If you can’t do that, refine the target outcome until you can. You can use a worksheet like this: -------------------------------------------------------------------------------------------------------------------------------------Project Name: __________________________________________ Date: _____________ Product Vision or Market Opportunity:_____________________________________________________ Scope of the Project: [ ] New Product [ ] Revamp of Entire Product [ ] Tweak Existing Feature

[ ] New Product Feature [ ] Revamp of Feature

Product Domain:

[ ] Website [ ] Web App [ ] Mobile App [ ] SMS App [ ] Hardware + Software [ ] Whatever it takes Other constraints: ________________________________________________________________ ___________________________________________________________________________________ Measurable Real-World Outcome Sought: __________________________________________________ How Could You Measure It: ____________________________________________________________ What is Success By this Metric: __________________________________________________________ Target Actor: [You complete this in the next step]__________________________________________ Target Action: [You complete this in the next step]__________________________________________ --------------------------------------------------------------------------------------------------------------------------------------

Page 10

A Toolkit for Designing for Behavior Change

Step 2: Generate Potential Actions and Evaluate them 1. Write down who you’re thinking of serving. That’s your actor. 2. Brainstorm at least 5 very different ideas about what action your actors can take that will meet the outcome. You can think about a. the obstacles actors currently face to achieving the outcome b. what needs to happen right before the outcome c. what’s unique about your company that you can use to achieve the outcome? d. observe users who do achieve the outcome, and see what they do. 3. Generate behavioral personas: for each action, right down what different subgroups of people are likely to respond differently to your appeals to act. For example: people who are obese are much less likely to respond to an appeal to run a marathon than very healthy, in shape people. 4. Evaluate your actions according to 4 criteria: a. Impact: How well will the action actually achieving the target outcome? b. Ease: How difficult is it for the user to take that action? c. Cost: How costly will it be for you to build a product supporting that action? d. Fit: does the action make sense for the company’s larger goals and culture? 5. Select the best action according to user, and company, needs using these 4 criteria. There’s no hardand-fast rule here – it depends on the constraints and priorities of the company. 6. Take a look at the behavioral personas again; can a single product really serve everyone? Identify which personas in particular the product will prioritize, to refine and clarify your target actor. You can use a worksheet like this: -------------------------------------------------------------------------------------------------------------------------------------Project Name: __________________________________________ Date: _____________ Target Outcome: [From previous step] ________________________________________________________ Action

Behavioral Personas in Audience

Impact

Ease

Cost

Fit

Summary

Personas to Target

Action1 Action2 Action3 Action4 Action5

--------------------------------------------------------------------------------------------------------------------------------------

Page 11

A Toolkit for Designing for Behavior Change

TWO| E XERCISES Conducting the Discover y Phase on an Existing Product The goal of this exercise is to document, and then clarify, the outcome, action, and actor for an existing product. 1. Pick a product that team members have experience with. Ideally, it should be a either: a. a product that they are currently working on and need greater clarity on, or b. a product that they worked on previously that suffered from indecision or disagreement about the product’s goals. 2. Draw a large grid on the butcher paper or whiteboard, like this: Original Product Description

New, Clarified Product Description

Outcome Actor Action 3. Fill out the “Original” column based on how the product was originally envisioned: a. Outcome: what did the product owner want the product to change about the world? b. Actor: who did the product seek to influence directly? c. Action: what did the product try to have the actor do, that would support the target outcome? 4. Fill out the “New” column based on how the team would re-envision the product, to make it more clearly defined and targeted for behavioral impact. a. Outcome: what is the specific real-world change that the product should try to accomplish? i. What would make the product owner happy and benefited the product’s users? ii. Avoid states of mind among users (emotions, awareness, education) – look instead to the physical, real-world changes that they should cause. iii. Make sure the outcome is measureable. b. Actor: who should the product influence for maximum impact? i. Who are the people, ii. Where are they, and iii. What are their defining characteristics (age, income, etc.)? c. Action: what physical action should the actor take because of the product? i. The action should clearly drive the target outcome. ii. The action should be clearly measureable. iii. Be specific about what the action is, and how many times the actor needs to do it. 5. At the bottom, jot down the behavioral personas – groups of users who are likely to respond differently to the product’s call to action. And, determine whether you can serve them all, or you need on focus on particular groups of users over others.

Page 12

A Toolkit for Designing for Behavior Change

TWO| E XERCISES Developing the Conceptual Design The goal of this exercise is to start developing a narrative of how the user progresses from their current state of inaction to actually taking the action. The final narrative can be a journey map, customer experience map, a flowchart, or a verbal description; in this exercise though, we’ll start with a simple series of steps. Stage 1: Preparation and Initial Sequence of Steps 1. Pick a product that team members have experience with. Ideally, it should be a either: a. a product that they are currently working on and aren’t sure what do to b. a product that they worked on previously that suffered from low uptake or usage 2. Take the target action that the product seeks to drive, and put it at the bottom of the butcher paper or whiteboard. 3. Write out each of the steps that a person needs to take before that target action occurs. a. Don’t think about the product yet – just the physical things the person needs to do to get from point A to point B. b. If, for example, the person wanted to get a skin cancer test, what would have to occur beforehand? He would need to go to the doctor. Before that, he would need to make an appointment with the doctor. Before that, find a doctor (dermatologist). Before that, starting look for a doctor, etc. The top of the page (and start of the sequence of steps) should be whatever the person is “normally” doing in their daily life when you are trying to drive this action. 4. Look over the list of steps for completeness. a. Are they are major actions missing? Add them in the sequence. b. Naturally, there will be multiple possible routes a person could take to get to the target action. Start with the most obvious one; we’ll refine it. (There’s another exercise to evaluate possible actions for feasibility, etc.). 5. Mark off which steps can occur within the product, and which ones occur outside of it. Stage 2: Structure and Refine the Action Itself 6. The sequence of steps the team has outlined isn’t set in stone. The best way to help someone take an action is to make the action easier for them to take (NOT to design a more compelling product that pushes them to take an action that’s hard). 7. Cheat. Look over the sequence of steps for things that aren’t necessary, and can be dropped. Then look for things that the product can automate or default, so that the user doesn’t need to do anything at all. 8. Tailor. Look over the sequence for ways to make the steps similar to, and building upon, other things that the user already does. 9. Simplify. Look over the sequence for ways to make the steps simpler and easier to do. Look for the Minimum Viable Action – the smallest, simplest thing that the user can do that still accomplished the target outcome. 10. If there isn’t a lot of time, the team can stop here. The next two stages are part of the longer, more detailed design process. Page 13

A Toolkit for Designing for Behavior Change

Stage 3: Construct the Environment. 11. The environment in which the user decides to act (or not!) has two parts: your product, and the rest of the physical environment that distracts or encourages them. 12. At each step, jot down notes about whether the user has the five preconditions for creating action. A Cue, a positive intuitive Reaction, benefits > costs in a conscious Evaluation, the Ability to act, and Time pressure, in order to then Execute the action (That’s C-r-e-a-t-e). 13. For the steps that occur within the product, what can the product do to fill in the gaps? Does it need to further motivate the user? (Part of the conscious Evaluation). Does it need to Cue the user to act? Etc. 14. For steps that don’t occur within the product, can the product recommend changes that the person makes to their own environment? Like Cues: setting an alarm clock, or putting a sticky note on the refrigerator? Stage 4: Prepare the Individual 15. What prior information does the individual need to take this action? Can the product supply all of the necessary logistical information so the person doesn’t need to plan out what to do? 16. How can the product (including the marketing of the product) make the person think about the action as a natural extension of who they already are? How can it make the action seem obvious?

Page 14

A Toolkit for Designing for Behavior Change

THREE | Q UICK R EFERENCE Illustrations of the Process

Page 15

A Toolkit for Designing for Behavior Change

Page 16