IA Summit: BJ Fogg
I am at the Information Architecture Summit, blogging the sessions I'm attending.
BJ Fogg is a professor of Computer Science at Stanford, and he will be telling us about persuasive technology.
Let me start by asking, how many people have been involved in a car
crash? (Most hands raise) How many have had a *pretty bad* car crash?
(still quite a few hands raise) I was in one pretty bad crash
once, and the thought that was on my mind that instant and is now
irrevocably burned into my brain was not my life flashing before my
eyes or anything like that. It was just one very simple thought:
"Control-Z!". I take this as convincing evidence that technology has changed me. Sometimes during conversations I feel like I want to reach for the Tivo remote - I want to go back a little bit.
The tools that we use, they do change us. But if you look back, it's
been that way for a long time. I dated a ballerina once. Her toes
wore all scrunched up. Her ballet shoes did that to her. So few
people can change so many people in the world. The few of us who are
designing those systems, we are affecting large numbers of people.
We're creating environments. How many hours a day are you doing email?
How much time do you spend in a browser. Four? Six? That's effectively
your environment. We are changing the way people's minds works.
I believe we are changing culture, too. Back in the day there were
shamans, etc., who dictated and shaped and ran things and people. We're
in the position where we're taking on that role. It's a little scary. I
don't think any of us went into this thinking we wanted to be the
shamans of the new world. We're designing the rituals of the
new world - that is, the best practices and assumptions of the culture.
My (semi-guilty) daily ritual: get up, turn on the computer, check
mail, look at the news online. (someone asks BJ to speak up) How far back shall I rewind? The functions of rituals are: 1. To reinforce a mindset or worldview. 2. To pass it along so it continues existing. Those of us who do this kind of things, along with Hollywood, are changing the world.
The Ctrl-Z and ballet toes are unintended side effects. What I want to
talk about today are planned effects. Often people don't think through
this. Two years ago I was invited to this wedding in a rich
spot north of Stanford. Three people with me. Just before we go in, one
of my companions asks, "How do you want to be introduced? Who do you
want to meet?" Another replies, "Oh, I want to be introduced as..." We
had planned effects. It turned out to be very effective. The people at
the wedding were media people, capitalists. One introduction led to
another: "oh, you need to talk to John..." I used to go
kayaking. As you're going through a rapid, you don't just drift. You
gotta paddle, it gets you through. I want to use this metaphor: are we
thinking about how our users paddle through their experiences? When you
plan the paddling, you reduce the side effects. When desigining a
website, paddling means saying things like, "We want visitors to think
this is the most credible source on cancer treatment". Drifting, on the
other hand, is thinking along the lines of "we wanna make it easy to
use, and, uhm, look good, too." In this room, you don't
want to be in the business of persuading people. Can you think of three
websites that don't have persuasion as a goal? The web has gone from
being this über-source of information to something that changes what
you think and what you do. Here's a little clip we prepared.
Bongo's planning a party. Bongo wants to know if the weather's going to
be good on Saturday. He turns on his computer. The computer wants Bongo
to update stuff from Microsoft. Bongo just wants the weather. Then he
gets a message from Britney Spears. And the computer says he needs to
backup. It wants to check for viruses. Bongo says no to every nag from
his computer. He finally manages to get the weather info. Bongo starts
to turn off the computer, and then the weather website interrupts -
"make me your homepage?", it asks. He hesitates, then says no. Bongo
feels poked, prodded and pestered by his technology. How much
time do we spend being interrupted rather than doing what we set out on
doing? I'd say more than 60% If we map the evolution out, 1980 would be
pretty low, then it would shoot up. I've thought about monitoring this
on large numbers of users' machines, but then thought, "If you present
something at a conference based on spyware they'll throw you out."
It's very different living in a small town. When you expect repeat
encounters, it changes the way you drive, all kinds of behaviors.
Let's talk about captology. I went to Stanford, I wondered what happens
where persuasion and computers overlap. After about 2 years of looking
for people who had worked on this, I concluded that there was very
little in that area. I said, let's give it a name - "captology".
Tech evolution is all about getting the darned thing to work. 1.
functionality; 2. entertainment; 3. ease of use; 4. networking; 5.
persuasion. Before the Web really took off, I was designing careful
experiments at Stanford. What we showed was that computers could make
people work longer, feel better about themselves, etc... I presented
this at a CHI conference. The reception was not very good. Some
researchers said "this is evil, we shouldn't be researching
this." But others said excitedly, "I'm in business. We should
talk". As I was wrapping up my book, I had become increasingly
become uncomfortable with the notion that computers can actually
control human behavior. I went into my lab, and I said, "Okay, no more
hiding. Our lab will now have a point of view." I got some pushback
from Ph.D. students and others in the lab - they were saying that
scientists don't have a point of view. But we do. Our point of view is
that computers do have good and bad effects. The bad: Large
organizations use tech to take away choices from individuals. The
good is when tech enhances choices and helps you achieve goals.
This is what happens when you install Quicken. [shows some cryptic
message in a yes/no window]. I'm a computer scientist, and I can't
understand it. The reaction is to say yes. This is the rubberstamping
strategy for influencing behavior: the design is done in such a way
that most users will say yes. Another strategy for influencing
behavior is the direct request & carrot: "Please Register - here
are arguments why you should do it". Piggybacking is putting ads on
your desktop that have nothing to do with what you were installing.
Quicken has evolved into persuasive technology. Applications start off
useful, they become richer, then persuasive, coaching you, with the
intent of modifying your behavior. There can be a good side to it. When
I do my taxes I like to be guided. Captology applies to many
domains. I'm most intrigued by personal technology. Politics. 4 years
ago we analysed candidates' websites. Fast-forward to 2004 there was a
significant improvement. I've been thinking about our
responsibility as computer scientists and teachers. We need to be
responsible. I don't think legislation works, I think we have to
educate large numbers of people. Bongo was a prototype. We've been
making those mini-commercials. I'll show you some other clips my students have done. Those are ads.
PAM ad: Personal fitness coach: carry it during the day; plug it into
the computer website loads automatically. MOnitor calories, etc. Plug
into an online community. Pam keeps you moving. PAM is a mobile device.
The complexity happens on the website. Health ad. Academics
each have theories, and they're not trying to fit together. I think
there's always been about 60 persuasion techniques. Mike ad.
Surveillance - If you really need behavior change, this one works. Meet
Mike. Mike monitors what applications you're running and what sites you
visit, so your boss can install Net Nanny. NN blocks some sites, saying
"Get back to work you lazy slob!". You'll have to play solitare with
real cards. ---- How do you design for persuasion? Most
important and easy is sign-up and profile, and invitation. So many
companies have no idea what the goal is for their website. Some
persuasion strategies with powerful impact: praise, persistence,
barrier reduction, immediate rewards, pain and fear, social influence,
stories (cause-effect), hope. Hope? I once set out on designing
a system that encourage people not to buy lottery tickets. But we
realized that for one dollar, they're buying hope. We dropped the
project. [SP: I rather think they're buying hopelessness]. Video
games have fire in them; they get many things right. Rewards. Constant
indicators and feedback relating to your increasing ability. Cause and
effect. The US army got a video game out that is effectively a piece of
training for the army. They got millions of downloads, that's
impressive. Classical conditioning is the Pavlov thing. Clicker
training. You click and feed. Then the click becomes as good as the
reward. Then comes operant conditioning. You can even train fish! If
you're being reinforced moment by moment, you'll keep on doing
it. Computers can train you just like we train dolphin, if they
can sense what you're doing and dish out a reward when you're doing the
target behavior. Periodic reinforcement is more effective than
systematic reinforcement. That can be really exciting, and really scary.
Let's look at sounds. People hate sounds that indicate that something
wrong is happening. There were inter-group differences within our
sample, though. You can target pleasant sounds to a pretty narrow
demographic. Fogg's Maxim for Credible Design: "To increase the
credibility impact of a website, find what elements your target
audience interprets most favorably and make those elements most
prominent." Captology blueprint: persuasion profiling. There's
this online retailer that is building a sophisticated map of user
behavior. They might use it against us. They will sell it too. Their
clients will include politicians. Why not? These are not churches. An
online bookseller may know more about my motivational psychology than
my brothers and sisters! Their minimum study has 60,000 users
in it. To a researcher that's very appealing. With that kind of sample
you can try many strategies, not think too much about it, and see what
sticks. What do we do with this? I want to start with a little
story. I come from a big family, seven siblings. At some point we
stopped buying gifts for everybody. Now we draw names and each one
makes a gift for one sibling or parent. A couple years ago my father
was supposed to get me a gift. His hobby is woodworking. He turned me
this bowl. "I still need to sand off the lacquer", he said, his
fingerprints were on it. I said, "You're not sanding them off! I want
to keep them!" In order to go in the right direction, we need
to look at how we live our lives. Methods matter. 1. Specialize. Find
something that you do better than anybody else in the world. The more
you specialize, the broader your impact. Force per area means more
impact. This is one of the mistakes I did early on, trying to spread
myself too think. 2. Take risks. I climbed the Kheops pyramid once. It
was risky and illegal and I'm afraid of heights, but I saw the sunrise
from the top and it was awesome. 3. Appreciate. In the last 3-4 years,
appreciation has become huge in my life. There's something about
appreciation that puts us right with the world. There's a company,
HeartMath, that studies the physiological effect of appreciation.
Apparently your heart and brain rhythms synchronize when you
appreciate. In almost every spiritual tradition, appreciation is
key. 4. Rebound. I've failed a lot. I have a lot of stories I
could tell about failure. Just get up and keep going. The world keeps
going. I like to think of myself as one of those toys that keep bobbing
back up. Let me wrap up by talking a bit about this antique
compass I found in a Japan store. It has old characters on it. For me
it is a metaphor about guidance. Where is my true north? When working
in an area that can be really good, or really bad, this matters.
Growing your freedom is what matters. I don't do things for money or
fame. My compasses help me find my direction. Ask yourself, what is
your true north? Ask others who know you well if you can't figure it
out. Community is very important. Who's going to join you in your
journey? My ancestors were pioneers, they constantly faced challenges,
but they had community. When they had conflicts they got over it and
moved on. I hope we can do the same. Any questions? PeterMe:
I work with a nonprofit whose goal is to "improve access to healthcare
in california." How to translate something broad and vague into
concrete form? Do you have experience with more mission-based
organizations, to translate that into specific behavior? BJF: Break it down into something that's concrete and measurable. Design for those pieces. Q.
"We are the new shamans." In this community I disagree. I think we are
the altarboys of the real priests, who are the technologists and the
programmers, which is where the bad things happen. BJF. I get
the feeling that this is going where engineering is becoming a
commodity and people working on user experience are becoming central. Q2
(same person). Maybe in new product development... Many times
information architects are brought in after the goals have been set. BJF.
Take heart in the the fact that the human experts are going to become
the people who specify what gets made. Maybe do some research on your
own to try to get a foot in the door in terms of setting direction. Q.
(Sandra) I really liked your comments on videogames to train behavior.
There's this game where you feed the starving supermodel, by throwing
food at her. You get more points if you can get all the cheesecake at
that wench. She starts out at like 70 pounds. Don't let her die - if
you get her to 180 you win. BJF. and there's a site in Switzerland called "Catch the sperm", with a condom-like thing. Q. Are you familiar with classification and framing as a way to influence people. BJF. the vocabulary, the framing, what's stated and what's not is very powerful. Q.
I'm worried about seeing Microsoft Persuasion Server 1.0 come out, five
years down the road. Can we train people to detect persuasion? BJF. This is a superb question. The best way we've found so far was what I showed you, educational videos. Q. Just very briefly, can you rap on the connections between persuasion and power, and ethics? (laughter) BJF.
The more persuasive you are, the more power you have. Politicians have
been using fear and "trust me" to manipulate the masses. This is a
troubling area and I'm trying to be candid about it. Q. I'll be speaking about persuasive taxonomies tomorrow morning. Q. I work at one of those large evil corporations (missed the rest) Q.
(Rashmi Sinha) In our profession, allegiance to the user is sacred. If
our role becomes about persuasion, we become marketing. BJF. If it's something that's against the user, I wouldn't do it. Q. (Rashmi) About the lottery example, what about highlighting the probability of never winning. BJF. Go for the lightest touch. If it comes across as persuasion, people will reject it. Q.
Conduits of persuasion. We are working in the electronic medium, mobile
mostly. There's paper too. I think there are certain thresholds. In the
paper world there's junkmail, very overt. Electronic spam is adaptive. BJF. Computers have the big advantage of interactivity. It's very different from bumper stickers and T-shirts.
 9:38:43 AM
|
