Five Ways to Train Your Visitors Like Lab Rats
As a psychology major, one of the first lab experiments I did was teach a lab rat some little tricks, like pulling a lever or standing on its hind legs.
That’s kind of cool, but how does it apply to you? Hey, wouldn’t you love to be able to control your visitors like that?
Maybe you don’t want them to pull a lever, and making them stand up is the last thing you want them to do, but what if you can get them to comment, click a link, or do anything else more often?
When the rat pulled the lever, or stood on its hind legs, I gave it a small pellet of food. The rat ate the morsel and then went back to work, pulling my lever for me.
As an offhand note – the rat’s name was Cain. For some reason, my girlfriend thought this was uncute and decided to rename it “Sparklefairy.†In retrospect, she realizes this is an absolutely ridiculous name.
Fortunately, when it comes to training, human beings haven’t evolved much beyond rats. We see examples of this constantly, and we can use it when creating a site design.
There are five “schedules†of reinforcement. All of them are good, but one, the fifth one, keeps people coming back for more, and more, and MORE…
It’s the kind of thing that causes people to get addicted, and who doesn’t want addicted visitors?
Give it to ‘em every time! (Constant Reinforcement)This is the simplest kind of reinforcement. Every time Cain did something that I liked, I gave him a treat. Easy and effective, and he learned his tasks quickly.
Unfortunately, this schedule is much more costly than any of the other methods and you run the risk of satiating your subjects (meaning they’ll be full and won’t pull your lever anymore, because they’re not hungry).
Examples of constant reinforcement are purely information-based websites such as Merriam-Webster or Wikipedia. When you search for information on these sites, you get a result that fills your need, and therefore no longer need to look.
Make ‘em wait! (Fixed Interval)
Sometimes I’d set it up so that Cain would have to wait one minute after he was rewarded before he could get more. No matter how many times he pressed the lever, nothing would happen. Once the minute passed, the first lever pull rewarded Cain, and then he’d have to wait again. Quickly, he learned that pulling the lever right after a reward wouldn’t give him more food, and so Cain stopped working for me for free. Boo, Cain, boo.
This strategy by itself isn’t very useful. Forcing your visitors to not act is counter-productive. However, in conjunction with other strategies, it can have a very positive effect. For example, Yahoo! Answers rewards you for visiting the site once a day, giving you a small incentive to check in and see what’s going on in the site, and maybe answering (or asking?) a few questions. Blingo.com allows you to search ten times a day with a chance of winning a prize each time.
Try and Try Again (Fixed Ratio)
With a fixed ratio setup, I made Cain pull the lever a certain number of times before he received food. Once he learned the setup, Cain would pull the lever slowly after each reward, and then more quickly as it came closer to the last pull – knowing you’re closer to your goal makes you work harder for it. On the other hand, knowing you’re far from your goal may make you put off working for it.
Many pyramid scheme sites employ a fixed ratio reward system – “Get five of your friends to sign up, and get a free iPod!†Many of these types of sites also use points as a midterm reward - a constant reinforcement - to achieve the periodic reinforcement of the end result (an iPod, XBox 360, etc.)
Is it ready yet? (Random Interval)
With this schedule, Cain didn’t get reinforced until a random amount of time had passed since his last morsel of food. This resulted in similar learning to fixed interval, although slightly faster.
Generally, I’d stay away from this kind of reinforcement, as either fixed interval or random ratio is more appropriate and effective for almost any occasion.
Gambling Addiction! (Random Ratio)
Here’s the one I mentioned before – the schedule that gets people addicted. And frankly, what’s better than addicted users? Casinos make money hand over fist. Cigarette companies are still posting a hefty profit even with massive taxes and legal costs. Alcohol is a booming industry in good times and bad. And there’s almost no chance your site will kill people (Digg hasn’t – yet…).
Getting a royal flush in poker doesn’t mean you won’t get a full house on the next hand, so you’ll still try just as hard to win the next, and the bite of the gambling bug is something we’ve all felt at one point or another. At the same time, getting twenty bad hands doesn’t mean you’ll get another bad hand either.
Random ratio is simple in that every lever pull or comment or click will have an equal chance of winning, and often times the system allows you to win far more than you lose. Cain learned the system, and he realized the more and the faster that he pulled the lever, the more food he got. And he started working hard.
Now if you can implement a reward system in which essentially every action has a chance of achieving a (big?) reward, your visitors will be much more likely to do that action in a frenzy. Mix that in with the fact that people overestimate small odds and overrate their chances of winning large prizes, you’ll be able to create a site that people won’t be able to tear themselves away from.
And that means more money for you.
Stumbling your Way to 17,209 Loyal Fans - MikeBogo on Marketing and Monetizing
wrote @ 1:46 pm on February 17, 2007:[…] How does Stumble do it? In premise, it’s basically channel-surfing the web, but the concept is so easy and unique it gets fans. And it’s like gambling, where you’re betting your time for interesting content (it follows the most addictive kind of reward schedule - read more about that here). […]