Top 10 Research Findings 2012 # 2

This is a series now…practically an institution!

#2 Removing Gamification from an Enterprise SNS

I still haven’t found the real strength to write a post on Gamification.  It is coming, soon.  Until then, you’re just going to have to take this information for what it’s worth.  This study reviews the implementation and focuses on the subsequent removal of a gamification construct from an organizations Social Networking Site (SNS).  The study was conducted at a large technology based organization (something like Oracle).  Within that organization, a SNS already existed (something like sharepoint).

The researchers and the organization implemented a gamification construct that awarded points and badges within the SNS to approximately ½ of the employees.  The other half of the employees interacted with the SNS without the gamification construct.  The gamification construct was designed in increase interaction with the site.  As such, the gamification awarded behaviors such as creating new posts, uploading pictures, commenting on posts, and commenting on pictures by awarding points. 

Badges were awarded based on accumulated points and were displayed on the user’s profile page.  A leaderboard of points accumulated was also implemented.  The study was phased, with the first phase of the study lasting 6 months.  During the first phase, in which ½ of the organization interacted with the gamified SNS, the ½ that did not have the gamification construct was unaware of its existence. 

During the second phase, everyone in the organization received the gamified version of the SNS.  The third phase started 4 months later and involved the complete removal of the gamification from all emplyees.

The results of this study were really interesting.  First, the gamification construct did significantly increase content contribution to the SNS site when introduced to each group, but that increase decayed over time.  Second, new users who could earn points did add more content over the short and long term, but the proportion of new users who contributed was the same for the gamification vs non gamification site users.  Third, there were two main types of comments observed:  terse and target of interest comments.  After the gamification construct was removed, the prevalence of terse comments subsided.  Finally, the removal of the gamification construct significantly impacted the contribution of data.

Now let’s break those results down.  Once it was introduced, the number of interactions with the site went up.  Essentially, the gamification construct was successful.  The organization designed it to increase interactions and it did that!  The increases in interaction decayed over time.  This is not surprising, the novelty of the points eventually wore off.   The design clearly did not focus on sustainability, and if it did, it failed in that aspect.

So the gamification was a success.  But…and there is a large but here, the increase in interactions was primarily in the terse comments section.  Terse comments were things like Hi, Cool, or Awesome.  Meaning the gamification did not increase significant or meaningful comments.  Since the gamification was not designed to do this, that is also not surprising.  The real surprise comes in the information about new users.  Between the two groups in phase one (gamified and non gamified) the users who interacted with the gamified site did interact more than those without the gamified site in the long and short term.  The proportion of new users, however, who contributed at all was not different between the sites.  Essentially, for those who were motivated by the gamification, they used it more often.  Those who were not motivated by the gamification did not. 

This draws an interesting implication for gamification and any game used organizationally.  Not everyone is going to be into it.  The real trick is finding something that most people will be into, and figuring out how to design it in a way that really targets the behaviors you want to see.  The goalposts have to be in sight the whole time.  If you want meaningful interactions, design it to motivate meaningful interactions.  If you want volume, design for volume.  Do Not jump the logic train and assume that all interaction is good interaction.  You have to be specific.  Very very specific.

Thom, J.; Miller, D. R.; DiMicco, J. (2012).  Removing gamification from an enterprise SNS.  Proceedings ACM Conference on Computers Supporting Collaborative Work.