LP Login

Think Big. Move Fast.

I’ve posted in the past in praise of Amy Jo Kim’s principles of game design and in particular, the way that social media sites can use keeping score to drive the user behavior that they want.

Last week Rajat Paharia, CEO of Bunchball, pointed me to an interesting NY Time’s article that gives further evidence that users will do more for points than they would otherwise. The article summarizes one case from a very interesting paper published in the Journal of Consumer Research in 2005.

Points can give people the illusion of advantage.

The article describes an experiment where students were asked to do two tasks; one for a reward of a vanilla icecream and the other, slightly harder, for a pistachio icecream. About a quarter of the group did the harder task, which approximately matched the proportion of the group that preferred pistachio to vanilla.

A second group was asked to do the same two tasks, but instead were rewarded with points; 60 points for the easier task and 100 points for the harder task. Students were told that people who ended up with 50-99 points would be given a vanilla icecream and people who ended up with 100 points would be given a pistachio icecream (no option to take vanilla instead). As such, the two reward systems were functionally identical.

The surprising result is that when points were introduced, over half of the group chose the harder task and got a pistachio icecream. Yet when polled, the same proportion of people preferred pistachio. So when points were introduced, many people chose to do the slightly harder task to get the reward that they liked less, just because they got more points!

It seems that points alone gave people some satisfaction. The rest of the paper goes on to extend the findings in various ways that are not covered in the NY Times article.

Points can give people the illusion of certainty.

One group was given the choice between $1000 now or a 50:50 chance of $6000 or nothing in one months time. 30% of people took the money now.

Another group was given the choice of 1000 shares (points) now, or 3000 shares in a month. Shares today were worth $1. Shares in a month had a 50:50 chance of being worth 0 or $2. Again, the two groups faced choices that were functionally equivalent. In this group, 61% chose to take their chances and wait a month.

The fact that they were holding shares (points) gave the group a greater feeling of certainty, even though the value of those shares in the future was variable. Perhaps this is why people with frequent flier miles don’t always rush to redeem them even when their airline is facing potential failure.


Points can give people the illusion of linearity.

A group of people were asked to listen to an increasingly loud and unpleasant noise through headphones for as long as they wanted to.

One group was given M&Ms for listening; 10 M&Ms for the first 10 seconds, 9 M&Ms for the next 10 seconds and so on.

The other group was given points for listing; 10 points for each 10 seconds. The first 10 points that they earned could be traded for 10 M&Ms, the next 10 points for 9 M&Ms and so on.

Again, people rewarded with points listened for longer on average (median ~105 seconds vs median ~75 seconds) than people rewarded with M&Ms directly, even though the rewards were functionally the same.

Some people focused more on the linear increase in points than the non linear increase in outcome/rewards, and this kept them motivated for longer. There are clear parallels here to the established practice on increasing the number of points before “leveling up” in games.

How can you use this?

I’ve only summarized the results here – its worth reading the whole paper and thinking about how you can use points to influence user behavior in your business.

12 Responses to Social Media: More on how points can be used to drive user behavior

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>