Creating_Persuasive_Technologies-Fogg1.pdf

Creating Persuasive Technologies:
An Eight-Step Design Process

BJ Fogg

Persuasive Technology Lab
Stanford University

captology.stanford.edu
www.bjfogg.com

[email protected]

ABSTRACT
This paper outlines eight steps to follow as best practices in the
early stages of persuasive technology design. The eight-step
process, drawn from demonstrated successes in industry practice,
begins with defining the persuasion goal to match a target
audience with an appropriate technology channel. Subsequent
steps include imitating successful examples of persuasive design,
performing rapid trials, measuring behavioral outcomes, and
building on small successes.

General Terms
Design, Measurement, Experimentation, Human Factors.

Keywords
Persuasion, design, persuasive technology, prototyping, iteration,
behavior change, captology, behavior model.

INTRODUCTION
Fifteen years ago, there were relatively few examples of
persuasive technologies in our lives. The web wasn’t ubiquitous,
and software wasn’t designed to change behaviors; it was focused
more on crunching data and boosting productivity. But today
persuasive technologies are ubiquitous; we are surrounded by
digital products designed to change what we think and do.
Persuasive technology experiences come to us through the web
(from commerce sites to social networking), video games (e.g.,
Wii Fit and Dance Dance Revolution), mobile phones (e.g., health
applications for iPhone and commercial texting services), and
specialized consumer electronic device, from “talking”
pedometers to bathroom scales that track body mass.

Increasingly, the living room TV and even automobiles are
channels for persuasive experiences. For instance, TiVo not only
suggests programs to watch but integrates Netflix and encourages

customers to make purchases on Amazon. As for automobiles, one
feature of the Toyota Prius is a miles-per-gallon meter that
motivates owners to adopt more eco-friendly driving habits.

Today those of us who are interested in the design and study of
persuasive technologies have a wealth of examples from which to
choose. The existence of so many successful examples changes
the study of persuasive technology in significant ways. We no
longer have to invent new persuasive solutions out of whole cloth.
Instead, we can focus on existing persuasive technology products
and techniques, varying those systems to understand the dynamics
and principles of persuasive design. In this way, we can learn most
rapidly about the psychology of persuasion and persuasive
technology by working with existing solutions.

That said, there still will be times, either for commercial purposes
or for our own academic research, when we want to create an
entirely new persuasive technology for which there is no good
prototype. This can be a challenge, given that that many people
have little or no experience in creating products with a persuasive
goal, and our emerging field does not yet have a systematic design
process. The lack of a well-defined process for designing
persuasive technology leads people to adapt methods from other
fields, such as usability engineering, or to make guesses as how to
define and develop their products. Neither approach is efficient.

To address this challenge, in this paper I draw on my 15 years of
experience in studying and creating persuasive technologies to
offer what I consider to be “best practices” for developing new
digital experiences that influence people. I share and explain an
eight-step process for creating successful persuasive technologies,
focused on early-stage design. These steps are based on a
combination of my academic research, the many project-based
courses I’ve taught at Stanford University, and my work in
industry designing and testing persuasive technology solutions.
My goal in writing this paper is to help academics and
practitioners get started down the most efficient and promising
path for creating persuasive technologies.

THE PROBLEM: BIG FAILURES
As I’ve reviewed a broad range of work from students and
companies, I’ve found that attempts to create persuasive
technologies often fail. One problem is that many projects are too
ambitious, and thus are set up for failure. For instance, a design
team might select a challenging behavior, such as smoking

Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, to republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
Persuasive’09, April 26-29, Claremont, California, USA.
Copyright © 2009 ACM ISBN 978-1-60558-376-1/09/04… $5.00

cessation, as the target. While everyone would agree that helping
people to quit smoking is a good idea, such a target is too
ambitious for design teams that have never before created a
persuasive technology. Stopping smoking—especially for those
who have a long-time habit—is the Mount Everest of human
behavior change. Just as no one would expect a novice to succeed
in reaching the summit, novice design teams should scale back
their ambitions and save the difficult behaviors, such as smoking
or weight loss, for later projects, after they have learned to succeed
in designing technologies targeted at more tractable behavior
changes. As with any other endeavor, acquiring skill in creating
persuasive technology takes practice. Most attempts will fail, but
experience increases one’s ability to create successful products.

INCREASING THE ODDS OF SUCCESS
The purpose of the eight steps I propose in this paper is to outline
a path to follow in designing persuasive technologies that will
increase the probability of success. The process starts with careful
thinking, then introduces small, simple tests to produce
measurable success. Once a design team finds success, albeit
small, only then should the team attempt to achieve larger, more
ambitious goals. This is the route that leads to growing success
rather than ongoing frustration.

In recent years, I’ve encouraged academics and industry players to
think small while they are starting their design work in persuasive
technology. I’ve advised people to be less ambitious on project
goals and focus first on achieving small successes.

The reaction I get is sometimes negative. This doesn’t surprise me.
Many people are attracted to persuasive design because they want
to do something big and important with technology. In addition, at
times people feel compelled by an outside force—a grant, for
instance, or pressure from a company leader—to set out on the
most ambitious course. In the face of long odds, novice design
teams may insist on tackling the hardest issues, such as
alcoholism, weight loss, smoking, or global warming, because of
such outside pressures. These attempts almost always fail, either
because the project never gets completed, or tests of the
technology show that it has no lasting effects on behavior.

Most of the persuasive systems we see at conferences or read
about online are the few projects that succeeded. In this context,
where only the successes are visible, it’s easy to forget that the
landscape of persuasive technology is riddled with the carcasses of
failed projects. In fact, I’m certain the failures vastly outnumber
the successes.

Failure isn’t always bad. In fact, best practice in persuasive design
includes expecting to fail. But I advocate failing fast. This means
investing only a few hours in a trial that might not work. In
contrast, projects that require months or years should not fail;
that’s a waste of time and money. Large projects will succeed
when built on a foundation of many small, measurable successes.

Learning how to succeed with persuasive design projects in the
early stages is the purpose of the steps I propose in this paper.

Before I go on, a few words about my bias: While I view
persuasion as either attitude or behavior change, in this paper I
focus on behavior change in persuasive technology. I believe that
changing behavior is what matters most in issues of politics,
health, environment and more. Fortunately, measuring behavior
change is getting easier with today’s technology.

THE EIGHT-STEP DESIGN PROCESS
The eight steps in the process of designing persuasive technology,
described below and outlined in Figure 1, are carried out mostly in
sequence. In some cases, two steps may be carried out in parallel;
at other times, the design team may back up a step and re-think or
re-try. The eight steps are not intended to be a rigid formula;
instead, the steps serve as milestones to make the design process
more effective. Varying the sequence of steps to suit the
circumstances is a valid part of the design process.

Step 1: Choose a simple behavior to target
The first step in designing a successful persuasive technology is to
select an appropriate behavior to target for change. The design
team should select the smallest, simplest behavior that matters.
Often this requires a team to reduce their big goal to a small,
seemingly tiny, objective.

For example, I worked with a large health care company whose
goal was to help people reduce their stress levels. As an objective
for a persuasive technology design, that goal is too vague and
ambitious. So we sought to reduce the big goal to something more
useful for our early design purposes. After brainstorming many
options, we decided to target the following desired behavior:
Stretch for 20 seconds each day when prompted.

Note that this smaller goal was so simple that anyone could
achieve it, and success was measurable. Of course, we realized
that getting people to stretch for 20 seconds wasn’t our final
objective. But this simple behavior was a good starting point for
the larger goal: reducing people’s overall stress level.

A large, vague goal can be broken down in two ways. Sometimes
a small goal can be an approximation of a larger objective; for
instance, stretching is a small solution for reducing stress. Other
times, a small goal can be a first step in achieving the larger goal.
For example, watching a short video online about Pap tests can be
the first step in persuading women at risk to actually get tested.

Achieving the small goal may have bigger effects than expected;
persuasion professionals have long understood that getting people
to do small things naturally leads to their adopting more ambitious
behaviors, even without a bigger intervention. For example, if the
big goal is to get people to be more environmentally friendly, the
small step of motivating them to change a single light bulb in their
home could alter how they view themselves, and over time they
are more likely to make other eco-friendly choices that are
consonant with this small step.

This first step in the process—choosing a simple behavior to
target—is the most important aspect of designing successful
persuasive technologies. But taking this step is more difficult than
it sounds. The best designers will advocate for simplicity, but
doing so takes courage. When working with a team or under
supervision, a designer risks looking timid when saying “no” to an
ambitious goal and proposing instead something small and simple.
In some situations, members of the design team may already have
their own pet ideas for what they want to build, so the team
collectively nods while each person adds an additional bell or
whistle to the user experience, complicating the project and
unwittingly setting the team up for failure.

Step 2: Choose a receptive audience
Step 2 in the persuasive design process involves choosing the right
audience for your intervention. Where designers have a choice
(i.e., the audience is not pre-determined by the project), I advocate

choosing the audience that is most likely to be receptive to the
targeted behavior change. The audience also should be familiar
with the technology channel (I will discuss channel in more detail
in Step 4).

The team may be tempted to design an intervention for the
toughest audience, such as helping compulsive gamblers to stop.
In my experience, this is a mistake. A related mistake is to design
the intervention for all users rather than a specific user type.
Neither approach works well.

The goal of the Steps 1 through 7 of the design process is to create
a digital product that reliably persuades someone—not everyone—
to adopt the target behavior. In Step 8, I discuss expanding the
audience, but this step should be carried out only after the

technology has been successfully tested on a more responsive
audience. Until then, I advocate choosing the easiest target
audience. For example, if the team is designing a technology to
persuade users to adopt better eating habits, they should select an
audience who has demonstrated a desire to improve their diets. If a
team wants to persuade people to adopt a daily exercise routine,
designers will increase their odds of success by focusing first on
people who already exercise once in a while.

Design teams have so many things to worry about when creating a
new persuasive technology that a resistant audience is not helpful.
In fact, choosing the wrong audience will almost certainly doom
the design project, especially in the early stages. As I will discuss
in later steps, once a design team has developed an intervention
that is working, they will be able to expand their target audience
and bring in users who are less receptive to the intervention.

The next consideration in choosing an audience is how familiar
people are with technology. I advocate choosing early adopters or
other adventurous souls as a target audience. I believe it’s a
mistake to target an audience that is afraid of computers or is just
beginning to use the technology channel for which the team is
building a persuasive technology, be it texting, social networking,
or interactive TV. The best audience for early projects consists of
those who enjoy using technology and trying new things.

In some cases, the first two steps of the design process might be
completed in reverse . Sometimes the audience will
determine the target behavior, rather than vice versa. For example,
a project to motivate teens to save money is likely to target a
different behavior (e.g., getting into the habit of saving) than a
project to persuade older adults to save (setting aside a specific
amount to ensure a secure retirement). So if work in Step 2 causes
a team to back up to Step 1, that’s okay; finding the right
combination of behavior and audience is vital to laying the
foundation for the subsequent steps in the design process.

Step 3: Find what prevents the target behavior
Once a design team has selected the appropriate behavior and
audience to target, it’s time to move on to Step 3. In this step the
team must determine what is preventing the audience from
performing the target behavior. For example, if children in first
grade aren’t brushing their teeth each morning, what is lacking?
As another example, if alumni aren’t donating to their alma mater,
why not?

The answers to such questions always fall into some combination
of the following three categories:

• lack of motivation

• lack of ability

• lack of a well-timed trigger to perform the behavior

In other words, in Step 3 the design team must pinpoint why
people aren’t performing the behavior. Is it because they are not
motivated to perform the behavior? Is it because they lack ability?
Or is it because they are not being triggered to perform the
behavior at the right time? Or is it a combination of the three
factors? The answers in Step 3 will determine the work required in
later steps, so a thorough examination at this stage is critical. (For
more details about this method, see www.BehaviorModel.org)

Consider, for example, a middle-class family living in the suburbs
of Los Angeles. Suppose that the family is not at all eco-friendly.
If the design team’s goal is to motivate the family to use eco-

Figure 1: Eight steps in early-stage persuasive design

friendly light bulbs, in Step 3 the team should explore why the
family is not already doing so.

The design team may find that the family has both the motivation
and ability to use eco-friendly bulbs, but they are lacking a trigger
to perform the behavior. Or, the team may discover that the family
lacks motivation: They don’t see the benefits of using eco-friendly
bulbs. Or perhaps the family lacks ability—they don’t know which
bulbs to purchase, or they can’t afford them, or they feel they
don’t have time to change out the traditional bulbs in their homes.

Technology interventions that require only a trigger are the easiest
to create and the most likely to succeed. For example, in the stress
reduction project described earlier, in which the “small goal” was
to persuade people to stretch for 20 seconds, participants in our
pilot needed only to be reminded (one type of trigger) to stretch.
We didn’t need to motivate them to stretch, or teach them how.

In practice, a persuasive technology solution will often require
more than simply triggering a desired behavior. Rather, the
solution must also boost motivation or facilitate the behavior, or
both. If the target audience lacks only motivation, the persuasive
design should focus on motivation. If ability is lacking, the
solution should facilitate the target behavior. One caution: If the
target audience is lacking both motivation and ability, the team
may want to back up and rethink the previous steps. Early-stage
teams usually struggle with the task of creating a persuasive
technology that simultaneously motivates and facilitates a target
behavior. Achieving both at the same time is hard. In such cases, I
suggest redefining the target behavior or the audience.

Step 4: Choose a familiar technology channel
Once a design team has identified what is preventing people from
adopting the target behavior, they can move on to Step 4: choosing
the best channel for the technology intervention. Which channel is
“best” usually depends on three factors: the target behavior, the
audience, and what is preventing the audience from adopting the
behavior—i.e., the first three steps in the design process. What this
means is that in most cases, the design team cannot select an
intervention channel– web, mobile phone, video game, or other—
until the first three pieces of the process have been completed.
(I’ll address exceptions to this in a moment.)

Today, we have an increasing number of technology channels for
persuasion: Web, software installed on personal computers,
mobile phone applications, texting on mobile phones, social
platforms like Facebook, online video, platform games, and so on.
The challenge for the design team is to choose among the range of
persuasion channels available, considering how well each channel
matches the target behavior. For example, if a design team is
creating an experience to motivate donations to a political party,
the team will likely need to use the web in their solution to enable
the financial transaction. If the target behavior is to share a
message with at least one friend, then the channel could be email,
online video, or social networks, because all of those channels
make sharing easy.

The next issue in channel selection is audience. A design team
must select a channel that is familiar to the target user. I’ve
watched teams expect their audience to learn a new channel (such
as texting or social networking) and simultaneously adopt a new
behavior. This approach almost never works. I have come to
believe that most people can change only one behavior at a time.
And the reality is that adopting a new technology is a behavior
change. It’s unrealistic for designers to think they can layer in

another behavior change, such as daily exercise, without
overwhelming their audience.

If a design team must use a channel that is unfamiliar to the
audience, there is a process for doing so, but it takes time. To use a
new channel to change behavior, design teams must first help the
audience to become familiar with the channel. For example, if
AARP wanted to motivate older adults to walk each morning,
using text messaging as the channel for triggering the behavior,
the AARP design team would first need to educate their audience
in how to use texting, which in the U.S. is not common for older
adults. Only after their audience was comfortable with texting
should the team introduce an intervention to promote walking.

Training people to use a new channel is difficult, which is why I
suggest avoiding it when possible. The easiest and fastest way to
progress with persuasive design projects is to select a channel the
audience already uses. In some cases, this may limit the
intervention to widely used channels, such as email, web sites, and
online video. In contrast, mobile texting, video games, and social
networking are not familiar to everyone today and may work well
only for particular types of audiences.

Finally, the design team must select a channel that addresses the
answer to the questions from Step 3: Why isn’t the audience
performing the target behavior? Is the problem a lack of
motivation, ability, triggering, or some combination? The answer
will help to guide channel selection. Some channels, including
online video, social networks, and video games, are effective at
increasing motivation. Other channels, such as installed software
and specialized devices, excel at making a behavior simpler
(which is functionally the same as increasing ability). And some
channels, such as text messaging and email, work well for
triggering behaviors.

As an example, consider again the family in Los Angeles that is
not behaving in eco-friendly ways. If the family is primarily
lacking motivation, the design team should consider channels that
leverage motivation, such as social networks and online video. If
ability is missing, the team could consider a web service that
makes the behavior easier, such as a guide showing where to
eco-friendly light bulbs. Online video also could be used to
enhance the target audience’s ability to perform behaviors, such as
a video that shows the process for replacing light bulbs, step by
step. If the family is lacking only a trigger to change their light
bulbs to more eco-friendly versions, then email or text messaging
may be the simplest solution.

Re- ing the First Four Steps
Usually, the first four steps in the persuasive technology design
process are performed in sequence. But in some cases a design
team will make an exception and carry out the steps in a different
. For example, if the design team works for a health
insurance company, the team might be assigned an audience for
intervention, such as sedentary older adults; the team may not
have a choice. But they might have lots of flexibility about the
channel. In contrast, a start-up company that provides mobile
services will not have flexibility in terms of the channel choice
(they will use mobile, of course), but they may be entirely flexible
on target behavior and audience. Finally, consider the case of a
graduate student designing persuasive technology. The student
may not be personally attached to a particular behavior, audience,
or channel. But the student might have a constraint, such as a
funding source that determines the behavior, an advisor who wants
work done with a specific audience, or a class project that focuses
on a channel, such as social networks.

Designers work from different starting points, with different
constraints. As a result, they may carry out the first steps in the
design process in a different . As noted earlier, this eight-step
process is a guideline, not a rigid mandate. Designers should apply
it to the unique circumstances of each project. That said,
whichever sequence the design team follows, the first four steps
should come before moving on to Step 5.

Step 5: Find relevant examples of persuasive
technology
In Step 5 of the design process, the team should search for
examples of successful persuasive technologies that are relevant to
their intervention, as defined in the previous steps. Suppose the
behavior of interest is persuading people to donate money to a
particular cause. The design team must find examples of
persuasive technology solutions that succeed in getting people to
make donations. If older adults are the target audience, the team
needs to find existing solutions that work for that audience. Or if
the chosen channel is video games, the team needs to study video
games that have successfully changed behaviors.

A design team won’t always know if a given persuasive
technology is successful, because companies generally don’t share
their conversion data with outsiders. So in Step 5, making
educated guesses is a good approach. For example, if a leading
Web 2.0 company is using reminder emails to trigger behavior—
for instance, driving customers to the company’s website—those
emails are likely persuading effectively, or the company would
probably not continue this approach. Similarly, if a leading social
networking site attempts to get people to invite more friends using
a certain widget, then that widget is likely achieving the
company’s objective. In other words, a design team should find
relevant examples from companies that are succeeding, because
their solutions for persuasive design, even on small issues, are
likely to be effective.

In searching for relevant examples of successful persuasive
technology, design teams will rarely find one example that
matches the precise behavior, audience, and channel the design
team has chosen for its project. And even if this parallel example
were to exist, the team would want to study other solutions to see
a range of options. Specifically, a design team should examine at
least nine examples in total: three that achieve a similar behavior,
three that reach a similar audience, and three that use the same
technology channel as the design team’s.

Teams can also learn to design for persuasion by imitating the
methods of experts who work in that domain. For example, if the
team’s intervention focuses on persuading donations, a team can
learn from studying the best practices of professionals who
persuade people to donate.

With relevant examples of persuasion in hand, a design team is
ready to move to Step 6.

Step 6: Imitate successful examples
The next step in the persuasive design process is to imitate what’s
working in the successful examples gathered in Step 5. Ten years
ago, persuasive technology was so new that design teams needed
to create novel solutions. Today, the landscape is different. Rather
than starting from scratch, a better, more reliable method is to
imitate what’s already working—on Facebook, Amazon, in video
games, and more—and adapt those successful approaches to the
target behavior and audience. With so many examples of
successful persuasive technologies, there is no need to reinvent the

wheel. Identifying and adapting successful technology examples
to the design project at hand is the fastest, surest way to create
effective persuasive technologies.

The design team should not be afraid of doing something that is
similar to what has already worked. In the later stages of the
design process (Step 8), the team will have many ways to be
unique. The opportunity for real innovation comes after laying a
solid foundation.

Besides selecting the right behavior, imitating successful examples
is perhaps the most important step thus far in creating persuasive
technologies. Nevertheless, I’ve found that teams sometimes resist
this step because the result feels derivative. Innovators often want
to be creative; they hope to craft something completely new.
That’s a mistake—at least if the goal is efficiency in creating a
persuasive technology that works. When learning new skills,
people succeed fastest by imitating success, and the design of
persuasive technologies is no exception.

Step 6 requires insight. When a design team examines a successful
example, the team must be able to identify the “secret sauce” – the
special ingredient that makes the example effective. The secret
sauce usually is not superficial design elements such as color or
typeface. So the team’s challenge is to evaluate the example from
a psychological perspective to discover the essence of its
persuasive power. Step 6 is easier when team members have a
background in psychology and good intuition about persuasion.

The questions from Step 3 are a good starting point for Step 6:
What is the successful example doing to achieve behavior change?
Is it motivating? Providing ability? Triggering the audience to
adopt a behavior? Often, an example …