BJ Fogg directs research and design at Stanford University’s
Persuasive Technology Lab
, and is pretty much The Don of captology – the study of how computers can be used to influence people’s behaviour.
We asked him a few questions about how internet marketers could be using persuasion techniques more effectively, as well as some of the more scary implications for individual web users.
In a nutshell, what is captology?
Captology is the study of how computers can change people’s beliefs and behaviours. By saying computers, that includes everything from websites to mobile phones to video games. Basically, how can machines shape humans’ behaviours.
Your research into captology included training your dog using clickers. What did you learn?
Yes, I used clicker training to train my German Shepherd – which was fun, but also a way for me to understand better how computers use reinforcement to change our behaviour. It was so interesting.
I went in with the premise that you can read about how reinforcement can shape animals’ behaviour, but you don’t really understand it until you do it, and that’s true.
One of the big surprises was how much timing matters. If the timing of the reinforcer is a fraction off, you reinforce the wrong thing – which is actually good for computers as they are very responsive and accurate. You can tell the computer that this is the exact moment to do the reinforcement.
Also, I learned how easily behaviour is shaped. I taught my dog how to pick up some toys and put them into a basket, from scratch, in ten minutes.
Can you relate that to the use of persuasion techniques on a website?
Sure – pick a target behaviour. Take the example of posting a photo to a site. You want your audience to post a photo to your site every day.
That’s a complex behaviour, so you need to devise a system of rewards for people first going to the website, then logging in, then looking at photos, then posting one photo, then the final behaviour which is posting one every day.
You think systematically about the behaviour and the steps that lead to that behaviour. But the trick isn’t so much coming up with the steps, but coming up with the rewards.
We did a series of studies into how audio can be rewarding, and ended up with 14,000 evaluations on 400 sounds. We came up with conclusions about what were the most rewarding sounds.
What were they?
We did the study on people in Japan and the US. We found that in both cultures, sounds that convey happiness or peace, such as a babbling brook or a child laughing, are very rewarding sounds. Sounds like alarm sounds are the most punishing.
If you go to Yackpack.com [an online audio messaging system for groups], which is my company outside Stanford, you hear a harp sound when you send a message. It’s a bit of a goofball but it’s a sound that people enjoy.
Can you give some examples of websites that are using captology very well?
There are a lot of them. In the fall, my course had 40-odd students and they went out and looked at Web 2.0 companies and captured in video what they were doing to achieve their target behaviours. If you go to captology.tv, those videos are there.
That exposed me to a lot of things that I hadn’t seen before. I’ve used Facebook but am not addicted to it, but there are a couple of videos that show why people come back time and again to the site.
Also, LinkedIn – they have worked very hard, I think, to fine tune that persuasion, and I think that’s why they have won instead of their competitors. Things like Digg are also interesting – why people put their time and energy into it. Why? Because there is some kind of social reward.
Video games are also very interesting. I’ve been looking at the game BrainAge, which is for the Nintendo DS, and looking at the persuasion strategy in that.
How do you think the big online retailers are doing?
There was a wave of it with retailers, but I think retailers have failed to innovate in terms of persuasive technology.
Amazon represented one of the early examples of persuasive technology. Reducing the complex behaviour of purchasing to a single click was brilliant, and it is doing some interesting things with recommendations, but they are less interesting than what has happened with social networks and Web 2.0 companies.
It is time retailers went to the next level in persuasive technology.
That doesn’t mean people buying things that they don’t want – it’s about helping people find things that meet needs in their lives, and it’s hard to think of anyone doing that right now.
What’s your view on the potential of mobile phones as a platform for persuasion?
The mobile phone, since it is with you so much and will eventually know so much about you, has the opportunity to persuade you to do those things that are going to be beneficial to you.
All the right pieces are there, but there is a colossal failure of cooperation, and it is going to take a while. The barriers aren’t technical, but more within and between organisations. To get an application on a mobile phone is very difficult.
Ironically, developing countries may end up quite far ahead of us in mobile applications as in some countries, they still have a monopoly. The system is faster than in a landscape where the carriers don’t cooperate. My lab has tried to work with carriers and you can’t get anything done. It’s so hard.
So for retailers, there is a big opportunity coming up that they need to prepare for now. These problems will end up being solved, as there is so much money at stake. It’s time to go out and look at how your brand can translate into a mobile experience.
We just had an event on mobile persuasion in February and the reception was great. We’ve also got a conference coming up at the end of April that is bringing together the biggest collection of persuasive technology experts ever.
How closely are the ethics of persuasion being looked at? Are there any guidelines being worked on that would, for example, oversee how gambling companies can use it?
Yeah, it definitely brings up big ethical questions. We are creating machines now that can control human behaviour better than other humans.
Where this is going is that machines are becoming better persuaders than humans, just like Big Blue became a better chess player than the world’s best chess player.
When you look at the implications of that, it’s pretty scary. Companies that know how to do this will become extraordinarily powerful, and if it is used to promote war or genocide, it could become very frightening.
We’ve done a little bit of work on this, and there will be a session on this at the conference. We want to sound the alarm and promote conversation around these issues.
When it comes to targeting people based on their behaviour, what are the privacy implications for individual users?
The Federal Trade Commission asked me to testify in the fall. I had ten minutes and three of those minutes were spent talking about how I think we are now developing what I call persuasion profiles.
They will refer to that profile when they try and make you an offer. For example, some people will be motivated by a 20% offer while others will be motivated by time.
If you understand what resonates with people, based on their profile, you will be able to change their behaviour and manipulate them.
Is anyone working on collecting this data across a network of third-party sites?
I think that is definitely coming. I can’t give any examples of anyone doing it across sites, but you can see how there is a market opportunity for a company, just like people that produce credit reports and sell them.
Can you say who’s already working on this type of stuff?
I probably shouldn’t comment on that.
What are the most common mistakes websites make when it comes to persuasion?
The number one mistake is complexity. One of the most important things is simplicity – making it easy for you to do what you want.
Companies tend to overestimate how motivated their customers are, and even an extra click reduces conversion. You have to ask hard questions about whether each step is necessary, and whether it needs to be on the page.
The other mistake is not being friendly enough. Companies feel like that, to be respectable, they have to talk in a language like they are wearing a suit. The days of that are over.
Most people using the web today are younger, and even if you are a boomer and in your fifties, you like to think of yourself as younger. So you don’t need to wear a suit from a linguistic standpoint.
Look at Flickr as an example – they language they use is very casual and they have won a huge following.
What are your plans for Yackpack?
We’re working on angel money and have some company contracts, and are looking for ways to distribute it widely to virtual teams – people that work from different office locations or from home. Anytime that people aren’t working in the same room, but have to work with each other throughout the day.
Our team is now a distributed team thanks to using Yackpack. The surprise is that our team is now more unified now we are working remotely, rather than in the office. As a psychologist, I predicted that would happen as in the office, you have all this weird stuff going on.
Find out more about persuasion on BJ’s
, or in his book:
‘Persuasive Technology: Using Computers to Change What We Think and Do’