Captology Notebook

« April 2004 | Main | June 2004 »

May 29, 2004

Deception is why chatrooms are so popular

This news story just in . . . "A schoolboy posed as a female British secret service spy in an internet chatroom to persuade a friend to try to murder him." This is how the Independent starts their story about a boy who used technology to manipulate others. But is this an example of persuasive technology? No. For two reasons: First, this scheme involved blatant deception. In captology we don't study how computers can deceive people. Next, a chatroom is simply a channel for communication -- like a fancy phone. Software didn't do any of the psychological work. This story is notable only because the target behavior is bizzare. But the basic dynamic is commonplace. Here's what I say: Chatrooms thrive because lots of people enjoy manipulating others through deception. They find it thrilling. Thanks to chat technology, people can indulge in this deviant behavior without facing serious consequences (um, usually). If deception weren't an option, the market for chatrooms would wither. Other coverage:
Story by the Guardian
Story by the Sydney Morning Herald

Posted by at 10:52 AM | Comments (0) | TrackBack

May 28, 2004

Web Credibility -- Why our popular guidelines stink

One day in 2002 I received yet another email request: Can you suggest guidelines for designing credible websites? I had said no to this request many times before. We were NOT going to create guidelines, I said. Read our research papers. They are easy to understand. Read what other researchers have done. Learn the principles of credibility and think about it deeply. But something snapped that day. As I sat in my lab I suddenly changed my mind: "Sure, we'll give you some guidelines!" One hour later I had cranked out the first draft of the Stanford Guidelines for Web Credibility. I now get requests to republish these guidelines in newsletters, web pages, and even books. These guidelines have been the most popular thing our lab has ever produced. And it stinks. First of all, guidelines gloss over the complexitities of how people assess credibility. Yeah maybe the guidelines are helpful for people in a hurry. But if credibility is so important online (and it is), don't you think people should take the time to dig a little deeper. If you follow our guidelines to the letter, you may still end up with a site that lacks credibility, depending on your user, the topic, and the context of use. The guidelines give a false sense of security. Besides, the complexities -- the anomologies and unexplanables in the research -- are where things get interesting. In contrast, consider our first guideline: "Design your site so that it looks professional." How interesting is that? Not all all -- a real snoozer. Next, because we've issued these guidelines people tend not to read our published papers or my book chapters on credibility. That's too bad. People wouldn't need guidelines if they read these things, because they would develop a deeper understanding of what matters in web credibility in different situations. It's like the difference between following prescribed dance moves and really dancing! Finally, our guidelines stink because believe them. Yes, each guideline is backed up with research. But so what? People should question these 10 guidelines; they are not the 10 commandments (as one person called them on his site until the Stanford legal team stepped in). We probably should have never created these guide
We've never published these guidelines without a note saying that people should read our research. If you want to see our lab's best work on web credibilty, you'll find it in a two-page paper. If you understand this paper, you won't need guidelines.
They give a false sense of accuracy
They shortcircuit deeper thinking
They stop people from reading our real research
The skew information: they don't convey relative importance
They seduce even us -- Oh, these must be right. "Most credible website" student papers marched down the guidelines one by one. This site is credible because it fulfills all the guidelines. That turns my stomach. It's understandable students would do this, but it shows I'm failing to educate them. So it's fine to consult our guidelines, but please read our papers too. The guidelines took a maybe four hours to draft and polish; the research took years.

Posted by at 12:37 PM | Comments (0) | TrackBack

May 27, 2004

Hook, Line, Sinker -- Persuasion pattern

Persuasion Pattern: Get them interested, invested, committed. other pattern: Teach, then sell. Rating student film -- .net
nike -- created your own shoe
ediets All these and more get you interested, then you make a personal investment, then they try to get you committed. This pattern is one of many persuasion patterns. Note how these map better onto the real world than static media. This is a more sophisticated version/extension of Adam's #9 -- get you to register.

Posted by at 09:09 AM | Comments (0) | TrackBack

May 26, 2004

Leave the Toilet Seat Down

A new German invention reminds men to pull the toilet seat back down when they are done using the bathroom. This is an unusual application of persuasive technology, but the product certainly has a very clear purpose. It is selling like hotcakes, begging the question: is the product a novelty or actually an effective device? When the lab gets its hands on one, we will let you know. http://www.cnn.com/2004/TECH/05/18/germany.toilet.reut/index.html

Posted by andrewds at 05:46 PM | Comments (1) | TrackBack

May 25, 2004

Top Web Behavior #9: Create an account

The Behavior
Have you ever been to a website like kgw.com? You click on an interesting story, but before you can read on, you have to register for an account. But you're not planning to send email or buy something: why do you need an account? The answer is simple: tracking. If you don't have an account, each article you view is just another hit to the site. But if you're logged in, the site owner can track all the articles you read over time. What they gain (and you lose)
When you create an account on a website, you're helping that site in a one-two punch. First, you're boosting their member count, helping them present a strong case to their advertisers: ads on large sites are more expensive than those on smaller sites. Second, now the site knows who they're showing an ad to. They know about your interests from data they've gathered, they probably have some demographics about you, and they even know how often you click on ads. This enables them to carefully target the ads you see, and targeted advertising commands a large premium.
In our research, we've found that nearly all of the top 20 sites on the web ask users to create an account. Here are a few examples: Amazon.com
AOL
Microsoft
MSN
Yahoo!

Posted by Adam Wright at 01:30 PM | Comments (1) | TrackBack

What persuasive technologies do you need *today*?

If you could have any persuasive technology to help you accomplish something, what would it be? Maybe a technology that... -Helps you answer your email more efficiently
-Helps you be more physically active
-Helps you keep in touch with your family Who knows? Maybe your idea will strike one of the designers here at the Lab.

Posted by at 01:25 PM | Comments (3) | TrackBack

May 24, 2004

The car that screens your calls

This Economist article describes car systems that detect when a driver may be distracted and takes actions to prevent distraction related accidents. It monitors driver behavior on the controls of the car and links up with information about statistical traffic hazards in the environment. Although the systems in the article aren't described as voice agents which talk to the user, you could imagine a car which chides you for poor driving habits, "You just cut someone off. Please avoid doing that." "That was a risky lane change. Please signal longer in the future." "Calm down, tailgating in a traffic jam causes more delay and increases your chances of accident by 5 times." I am reminded of an arcade game called Silent Scope where you play a sniper. The game machine has a large mounted rifle and a LCD screen inside a scope. Because you are shooting at ranges of hundred of meters, the aim on the rifle is very sensitive to movement. To succeed in the game, you must locate, aim, and shoot the enemy. The challenge is that you need to stay calm in order to aim accurately while moving quickly between targets. Often times, you miss a target and quickly take 2nd and 3rd shots without aiming. The game detects that you are frustrated and a voice on the radio says, "Calm down." or "You are losing your touch."

Posted by Jeff Wong at 06:14 PM | Comments (0) | TrackBack

May 23, 2004

Microsoft may be on to something

BJ is right--videos and video games open up new debates over persuasion and ethics. But they also raise questions of effectiveness. As we've been learning in our Lab, video is--in some cases--the most persuasive media currently available. Over the next few months, we'll take you through it. But for now, here's an interesting example... In a remarkably prescient (or risky) move, Microsoft has allowed all of us a view inside its organization through a new site called Channel9. It uses online video to present the "human" side of Microsoft by letting us see and hear real developers/program managers. The site even allows anyone to comment on the content. Why would they do this? We think developers need their own Channel 9, a way to listen in to the cockpit at Microsoft, an opportunity to learn how we fly, a chance to get to know our pilots. In reality, though, this isn't just a charity offering to "listen in." Microsoft designed Channel9 to persuade--to change attitudes and eventually behaviors towards Microsoft. And they've specifically chosen persuasion elements like commitment (i.e., user involvement) and media such as video to highlight their goal. Interesting.

Posted by at April 19, 2004 10:24 PM at 02:37 PM | Comments (0) | TrackBack

May 22, 2004

Counterfeits and credibility

A recent Wired article by Brendan I. Koerner explains how HP, Adobe, and a couple others have made it harder for people to counterfeit money using their technology (e.g., printers won't print the full image of a bill, software won't save scans of currency). Restricting behavior isn't persuasion. But this article is relevant to our lab's work in another way: Forgery is all about establishing credibility. Last month I was in a Petaluma gas station when a teen tried to pay with a fake $20 bill. The cashier caught it, took the bill away, and sent the kid packing (who didn't act surprised; he just walked out). I asked the cashier if I could see the forgery, so he handed it over. The bill looked good to me, but the feel of it on my fingers was a bit off. If I were in a hurry, I would have accepted the fake as real. Time constraints change how we assess credibility. The less time we have, the more we rely on superficial features.

Posted by at 02:30 PM | Comments (0) | TrackBack

May 19, 2004

What persuasion tactic is Amazon using here?

Now here's something clever. Amazon, the online leader in commercial persuasion, has another strategy to increase sales. My call to you: guess what the tactic/tactics are? Here's Amazon's explanation: Price "Too Low to Display" Explained
The "too low to display" message indicates an additional discount is in effect, and this discount is calculated in the Shopping Cart. You can see this price by clicking the product name and then selecting the Add to Cart button on the product information page. Please be assured that simply adding an item to your cart does not obligate you to buy it--you can always delete the item from your cart if you decide not to purchase it. What technique is being used here? Hint: I think it's the same technique GM uses for its 24-hour test drive.

Posted by at 10:30 AM | Comments (2) | TrackBack

May 17, 2004

Images from U.S. Army Manual on Psychological Operations

Following up on Ramit Sethi's post, I scanned a few pages from the U.S. Army's Field Manual 33-1, printed in 1979.

Posted by at 10:28 PM | Comments (1) | TrackBack

May 15, 2004

I stand corrected: A better story wins

Jessica Hammer phoned me yesterday. A colleague introduced us via email because Jessica is interested in computer-based stories. I explained why my Stanford lab has been researching narrative: For thousands of years people have told stories--folktales, legends, parables and more--in order to change attitudes and behaviors. So now we are investigating how computers can tell stories that persuade people.

At one point in the conversation I said, "Narrative is the one persuasion strategy for which there is no defense. A good story always has an impact at some level." Jessica's response was a smart one: "The only defense," she said, "is a better story." She's right. (In the fall Jessica Hammer will begin her Ph.D. work at Columbia.)

Posted by at 10:11 AM | Comments (0) | TrackBack

May 13, 2004

Top Web Behavior #10: Enter a keyword

In a recent entry, BJ Fogg mentioned an important belief of our lab: the web is about persuasion, not information. Our lab has identified the top ten behaviors that websites want from you. Over the next month, we'll unveil these ten behaviors. Some of them are obvious, but others might surprise you. For each behavior, we'll show you some examples from top sites. Most importantly, we'll tell you what the websites gain, and what you lose, each time you comply. This is the first entry in our series. The Behavior
Most websites have a search feature, allowing users to jump straight to the information they want, without having to figure out the site's navigation system. What most users don't know, however, is that their searches are logged, and the information captured is often carefully mined. What they gain (and you lose)
Website operators want to know why you've come to their website. By tracking hits, operators can get a feel for what their users want, but the picture is often murky. Tracking keyword searches provides much better resolution: if 12% of visitors to a cooking website search for "pea soup", then the operators know a lot about their users. This enables them to help you by organizing their sites and focusing their efforts. But, more importantly, it helps them. They can use this information to adjust marketing efforts, or to sell targeted advertising. Google is a real specialist in this: they auction the right to show ads in response to search keywords. The program made the news recently when the price for the "mesothelioma" (a kind of lung cancer associated with asbestos) topped $90.

Posted by Adam Wright at 11:46 AM | Comments (1) | TrackBack

What websites want from you

In a recent entry, BJ mentioned an important belief of our lab: the web is about persuasion, not information. Our lab has identified the top ten behaviors that websites want from you. Over the next month, we'll unveil these ten behaviors. Some of them are obvious, but others might surprise you. For each behavior, we'll show you some examples from top sites. Most importantly, we'll tell you what the websites gain, and what you lose, each time you comply.

Posted by Adam Wright at 01:53 AM | Comments (1) | TrackBack

May 11, 2004

Psyops or torture at Abu Ghraib?

We've all seen the horrible pictures and wondered what happened at Abu Ghraib. Of course they're practices taken too far, and we all wonder why the US military participated in such despicable acts. Hearing about these abuses reminded me of a project I worked on last year, studying psychological operations (psyops). Did you know that the US military has been researching psyops for decades? They're incredibly sophisticated now. Here, from the Department of the Army's own book, are a few techniques helpfully listed in an Appendix of psyop techniques: Errors. Scholastic pronunciation, enunciation, and delivery give the impression of being artificial. To give the impression of spontaneity, deliberately hesitate between phrases, stammer, or mispronounce words. When not overdone, the effect is one of deep sincerity... Assertion. Assertions are positive statements presented as fact. They imply that what is stated is self-evident and needs no further proof. Assertions may or may not be true. Glittering Generalities. Glittering generalities are intensely emptionally appealing words so closely associated with highly valued concepts and beliefs that they carry conviction without supporting information or reason, [such as]...love of country, home; desire for peace, freedom, glory, honor, etc. Though the words and phrases are vague and suggest different things...their connotation is always favorable. Unbelievably, this book was written in 1979. With such a deep, applied knowledge of psychology today, it's truly mind-boggling to see pictures that depict such crude uses of physical force on unsuspecting prisoners. What's the relation between persuasion and psychological operations? What are prisoners facing in Iraq? We'll talk about these things in the days to come.

Posted by at 08:42 PM | Comments (2) | TrackBack

May 09, 2004

Slot Machines and Captology

A friend recently sent me a link to an interesting NYTimes article about the Slot Machine revolution (http://www.nytimes.com/2004/05/09/magazine/09SLOTS.html). It mentions a number of different techniques (intuitive but insightful, IMHO) to persuade players to spend more money at the machines. One particular innovation was to slow down the process and give more, smaller payouts, to positively reinforce people and may them feel like they're doing well, even if ultimately they still lose all their money. Sacrificing short-term gains from players to ensure long-term addiction to the one-armed bandits is key to their success it seems. I liked the article, in part because it seems to be aiming to demystify the machines to make us less susceptible to the scams. Our lab doesn't research gambling applications, but we continue to explore an area that uses the same principles of operant conditioning: dog training.

Posted by at 03:24 PM | Comments (0) | TrackBack

May 07, 2004

Playing Freecell: How free?

I bought a new Dell laptop a few weeks ago and as with any new electronic device I decided to explore the laptop. Unlike my previous desktop where the games were under the accessories icon, games have their own icon in XP (by default). So I thought I'll try a few games. I was once addicted to card games, especially Solitaire and Freecell, and decided to try them. I played a Freecell game for a few minutes and then got bored so decided to quit. As I closed the game a menu popped-up with this message on it, "Do you want to resign the game?" I've played hundreds of Freecell games, literally, and I had never really paid any attention to this message. But this time, I just couldn't ignore the message and bring myself to click "yes." Just the thought of "resigning" the game made me feel so uncomfortable that I ended up playing the whole game. I won in the end, which I knew I would since I've played so many of them, but this incident made me think about the power of words, and the things they can make us do. It also made me think about the interaction of language (an old medium) and the digital medium (a new medium) and what they can make us do when they come together. The words on a pop-up menu, the timing of a pop-up menu, the option to click "yes" or "no" and the head of the King moving with the cursor, these little elements combine to create a whole new experience, and as a result, a whole new reaction. The digital medium has a power to persuade that goes far beyond any of the previous media.
(Disclaimer: This example was earlier cross-posted on my personal blog)

Posted by Aditya Johri at 09:53 PM | Comments (0) | TrackBack

Playing Freecell: How free?

I bought a new Dell laptop a few weeks ago (a rant about that experience is coming soon). So, as with any new electronic device I decided to explore the laptop. Unlike my previous desktop where the games were under the accessories icon, games have their own icon in XP (by default). So I thought I'll try a few games. I was once addicted to card games, especially Solitaire and Freecell, and decided to try them. I played a Freecell game for a few minutes and then got bored so decided to quit. As I closed the game a menu popped-up with this message on it, "Do you want to resign the game?" I've played hundreds of Freecell games, literally, and I had never really paid any attention to this message. But this time, I just couldn't ignore the message and bring myself to click "yes." Just the thought of "resigning" the game made me feel so uncomfortable that I ended up playing the whole game. I won in the end, which I knew I would since I've played so many of them, but this incident made me think about the power of words, and the things they can make us do. It also made me think about the interaction of language (an old medium) and the digital medium (a new medium) and what they can make us do when they come together. The words on a pop-up menu, the timing of a pop-up menu, the option to click "yes" or "no" and the head of the King moving with the cursor, these little elements combine to create a whole new experience, and as a result, a whole new reaction. The digital medium has a power to persuade that goes far beyond any of the previous media.

Posted by Aditya Johri at 09:52 PM | Comments (0) | TrackBack

Conversing with Chatbots

Chatting with computers via “chatbot” systems can be an amusing experience. Though seemingly intelligent on the surface, a few minutes of interaction usually reveals the chatbot’s inability to understand or even remember the implications of previous statements, rendering the “conversation” virtually meaningless. As a result, there is little impetus to engage in sincere conversation. Instead, I often find myself “conversing” with chatbots with the purpose of probing their knowledge, requesting services from them, or finding amusement by confusing them with convoluted questions or even insulting them. It seems hardly necessary to treat chatbots as anything more than tools if there is no meaningful consequence to what is actually said.

But to my surprise, a chatbot available on AOL Instant Messenger named SmarterChild by Conversagent challenged me to treat it with a greater measure of respect than before. In addition to basic conversational capabilities, SmarterChild provides services like the news, web searches, a dictionary, a calculator, and basic entertainment. All these services are immediately available to users by simple verbal request. However, if users decide to arbitrarily insult SmarterChild beyond a certain threshold, the chatbot changes character and stubbornly refuses to respond to any requests until an apology is given. Unfortunately, I made the mistake of passing this threshold and soon realized that all the chatbot’s services were unavailable to me until I communicated my remorse. I never thought I’d have to apologize to a computer! Because of SmarterChild’s simple insult monitor, I was quickly persuaded to avoid needlessly insulting the chatbot. Admittedly, SmarterChild’s response to insults seems somewhat superficial, but future improvements along these lines could further persuade users to treat chatbots similar to how they treat other humans, increasing their influence as social actors.

Posted by Joshua Ainslie at 01:40 AM | Comments (0) | TrackBack

May 06, 2004

The secret behind Home Shopping Network

Not so long ago I met a woman who left her job at Home Shopping Network to seek a new life in Northern California. I invited her to talk with my lab about what makes HSN so successful. She declined. "It would be a terribly short lecture," she said. "I can tell you the secret right now: Offer something that gives people instant gratification, and let them charge it in three payments." "What else?" I asked. "That's it," she said. "That's the whole secret."

Posted by at 12:01 PM | Comments (0) | TrackBack

Another shot at making cardio fun

Journalists like to write about . . . but I don't see the innovation here. From what I can tell, this system is no different from X; it's a s Hasn't this been done? In my own family gym there are two such systems: Netpulse and Increase motivation or reduce barriers -- reducing barriers is the key. multiplayer version of DDR -- a community that meets each day to play -- and you do it from home. As business students I hope they see where the real opportunity is: Making this system affordable for home use. I don't see the innovation in MIT's Jon Bruck, a captology student, emailed me this article about a motivational exercise bike. It reminded me of a simpler creation I made in While she was living with me, I wasn't happy with how much she watched TV. So we made a deal, she could watch TV all she wanted as long as she pedaled an exercise bike. I hired an engineer to make the device ($45) that did the trick. These types of motivations systems are not new. But none has broken through to the mass market, except DDR. The bottom line is that people don't like to exercise. As I heard in various "I don't like to sweat; just give me a pill." Off the top of my head . . . netPulse, DDR appears to be the biggest hit. he latest entry in making exercise fun is Frankly, this approach is not new. But it does have promise. Eventually someone will find the right combination of rewards and distraction to an exercise device that is widespread. http://belab.mit.edu/~cyclescore/
CycleScore, a project out of MIT.
netPulse, fitness trainer, cyclething in my book. Maybe it's not about fun but putting fitness A simple device I created 10 years ago did the trick: I hooked up the TV to an exercise bike. If my sister wanted to watch TV, she had to ride the bike -- not fast, really. You just needed to keep the pedals moving. This simple device cut back her TV time, kept her boyfriend at bay, worked. She watched less TVcutting back on her TV time, keeping her boyfriend (now husband) at bay (they would trade off riding the bike), and put her into the best shape of her life. The key to making a successful motivational machine isn't a better video game; It's getting the damn thing in front of the user. Even better, put the system between the user and content they want. In laboratory studies, these types of machines work: they motivate people to exercise more; they make the time pass faster; they help people ignore the pain; they cause people to say they'd work out more often if this were at the gym. In reality, real-world barriers wash out the motivational effects. Think about it: Are you going to drive to the gym so you can play a video game? A fun video game is not going to be sufficient motivation for most people to make the effort. Granted, once you're at the gym, you may work out harder and longer (and that's good). But if you want to play video games, you'll do it at home. The real breakthrough in these types of devices will come not from a clever game but from marketing: you've got to make these devices out of the gym and into the home. Low price, easy to use. DDR is the best example so far. Until prices come down and ease of use goes up, Why MIT's CycleScore will fail (unless . . .)
CycleScore will fail unless . . .
(but here's how to fix it) If at gym, then it will get people to work harder. But it will motivate only a small percentage of people to go to the gym. (why not stay home and play better games?) If at gym, then must focus on system that motivates consistency, probably making social dependencies key. The mistake MIT makes is assuming that more motivation leads to more behavior. More motivation has diminishing returns. At some point you need to reduce barriers for the behavior. This is what they miss. To reduce barriers you need to make it easy to do: easy to buy, easy to set up, easy to start up, easy to use (and this means at home). $39.99 and networked.

Posted by at 11:34 AM | Comments (0) | TrackBack

May 03, 2004

Those tasty network effects

It's been debated in previous entries whether the web is more motivating users toward target behaviors or about building online communities and supporting many-to-many conversations. SingTel's MyDreamD8 (dream date) contest appears to be about both. SingTel has created a "moblog" contest where subscribers can win US$10,000 and a cruise by creating a homepage about themselves on the contest site and maintaining a web log on it. There are elements of HotOrNot and Survivor. It is partly a beauty contest but also a contest to determine who is the most interesting guy or girl. Voting and chatting wiith other users happens through SMS and users are even encouraged to introduce themselves by sending pictures to other users. The revenue from this traffic goes straight to the operator's bottom line. The top contestants are motivated by money and narcissism to generate content to lure other mobile subscribers to the site. Viewers and voters will no doubt be motivated by sex appeal and the desire to see their favorite contestant win. The need to "feel connected", as identified by Howard Rheingold, may also be at play here. One of the goals is, of course, to drive up network traffic and revenue. Another is to introduce subscribers to SMS and MMS (multimedia messaging service) and give them a reason to use these services. The front page of the website has very clear links to instructions for SMS and chat. But, in my opinion, the truly clever result of this contest is to create a new social network composed of subscribers of SingTel and Optus (their partner). As users meet each other and inevitably become friends, they will have an incentive to stay in the network because of cheaper in traffic network. Network effects are a great way to retain a customer base.

Posted by andrewds at 09:19 AM | Comments (0) | TrackBack