Knowing your user is key--Part 2: Motivation

A bit of a lag in posts--I was vacationing in Maine and although I was thinking about where to go next on the blog post, I didn't actually get around to writing anything. I know, slacker.

Picking up on the theme of motivation, let's talk a little bit more about specific things that motivate people.  Here's the funny thing about intrinsic motivation versus extrinsic motivation.  If you aren't careful, you can take something that people find intrinsically motivating (that is, they just think it's fun) and you can suck the fun out of it by adding an extrinsic motivation like a reward.  There are a variety of studies on this going back to the 1970's (see classic articles like this one by E L Deci or by Lepper & Greene).  Here's the classic example.  You take a group of children who have been identified as liking to read and you divide them into two groups.  One group gets told the more you read, the more we will give you a reward, like $1 per book.  The other group doesn't get told this (or in another study, only gets an unexpected reward later).  After a while, you measure the number of books the kids are reading and you see that the kids you gave a buck a book are reading LESS than the kids who didn't get anything or didn't expect anything.  You took something they liked to do and attached a reward and they stopped doing it.  The suggestion is that if you take something like reading, which they were doing for pleasure and make it about doing it for money, they stop enjoying reading for its own sake and when the money isn't something they care about anymore, they stop reading.  On the other hand, if you take a third group of kids who don't like reading and give them a reward, their reading behavior increases.  They aren't doing it for the fun of it, but you can give them an incentive that will increase the amount they read. 

The story of course, is never quite that simple and since the 1970's, psychologists and economists have come to realize it's not quite that easy.  There are a lot of nuances to when extrinsic rewards work or don't.  If you follow Deci and Ryan's 1985 cognitive evaluation theory, anything that leads to a perception of self-determination or of competence will increase intrinsic motivation, while anything that decrease those perceptions will also decrease intrinsic motivation.  Eisenberger, Pierce and Cameron (1999) did a nice meta-analysis of the studies on the effects of reward on motivation and concluded that rewards for things that require minimal performance or are not well defined decrease intrinsic motivation, probably because they convey to the person that the task is trivial.  However, if the task requires high performance, it conveys that it is important and rewards actually increase intrinsic motivation.  Basically, if you reward trivial tasks, people catch on, and they lose any intrinsic motivation they might have had.  That could have a real implication in gamification.  If you reward people for pointless things, they lose interest and can even be really turned off by it.  My own personal example of where pointless badges are an epic fail is the Google News Badges--Techcrunch summed it up nicely in this piece.  (You can read up on a lot of Deci's work here if you want to see how that theory has morphed over time into the self-determination theory.) 

So where does this lead?  In part, I think to the conclusion that if you are not clear what gets you rewarded, you won't increase that behavior and if you reward trivial tasks, you can decrease intrinsic motivation.  People are driven by a lot of things, and it's important to understand what motivates them.  But if you want reward to work, you need to clearly tie it to the behavior you want and you need to expect high performance.  I'll go more into drive theory on the next post, but what we're building up to is how all of this plays into the use of gamification in enterprise software.  Lots of psychology behind how we get to this point!


Finally someone is talking about this in a gamification context. Thank-you for pointing out that it is complex, nuanced, and easy to get wrong (with potentially damaging consequences).

I want to mention Dan Pink's "Drive" as a user-friendly intro to the underlying research of Self-Determination Theory. He gives some simple guidelines (over-simplified, but good for avoiding the danger zones) on the use of extrinsic motivators for things that can be intrinsically rewarding: do NOT use rewards-for-behavior, in the form of an if/then (if you do X, you will get reward Y), though this is the most common form of gamification today -- the pure operant conditioning of reinforcing the "behavior we want."

However, a reward that functions as a recognition of significant achievement (not the faux-achievements of most gamification, but real performance like earning your next martial arts belt) is far less risky. And of course giving extrinsic rewards for behavior that will NEVER be intrinsically rewarding (enjoyable for its own sake) carries little risk of DEmotivation since the behavior is-- and probably never will be--intrinsically motivating on its own. Which is why gamification of exercise, health compliance, etc. is much more likely to be a safe and effective use of typical (shallow) gamification to increase behavior.

Where it goes so horribly wrong is when a company wants to use typical extrinsic rewards (externally-regulated reinforcers) to increase a behavior that is -- or needs to be -- intrinsically motivated (or at least *integrated*, the part of the motivation continuum closest to intrinsic motivation). Sharing with co-workers, social engagement, etc. are all areas where typical use of external regulation risks the HUGE problem Deci, Ryan, and others (Amabile, recently) describe. Of course these operant conditioning reinforcers DO often create a lot of "engagement" and activity, so they can look quite successful, while simultaneously reducing the very motivation the company is hoping to achieve, for the long haul. Sustainable, intrinsic engagement bears little resemblance to engagement around a reward structure, though the difference might not be obvious at first, and only appear when things begin to deteriorate in an unrecoverable way.

I think there are few who take this research seriously (despite Pink's NYTimes bestseller about it, and an uber-popular TED talk) because it is so counter-intuitive. Most people assume that rewards for behavior are at best a massive win for everyone, and at WORST benign. Yet hundreds of studies continue to show the bizarre yet reliably reproduced DEmotivation that can happen under these very specific circumstances. Unfortunately, those specific circumstances happen quite a lot in business, especially in areas around social media use and deepening authentic engagement.

Posted by Kathy Sierra on August 03, 2012 at 04:56 PM MDT #

Thanks Kathy, glad you liked the post--and you have actually predicted exactly where I'm going next--this post was originally titled Motivation, Drive and Game Mechanics, but it got too long as it was, so I'm making those separate pieces.

Posted by Erika Noll Webb on August 03, 2012 at 05:00 PM MDT #

One more thing: your post appears to be about finding the *appropriate* use of rewards, including attention to what, precisely, is being rewarded. But this is where game design and gamification part ways. Game design is not about rewards *at all*, though they are often found as trappings and surface mechanics in video games -- a subset of all games. Every day, thousands of people in the world play one of the most popular games on the planet (and one of the oldest) -- the game of Go. Though not popular in the US, pros in this ancient game live like rock stars in some countries, and the online Go servers are extremely active 24/7. And very few players earn-- or even consider-- the use of external rewards in any form.

Virtually no strong Go player is doing it for a place on a leaderboard or FOR a higher ranking. What a higher ranking in th game enables (just as with chess or a martial art) is the ability to have even more intrinsically rewarding experiences as you are the able to play at a higher level, with others of a greater strength, so the challenges increase in a deeply meaningful, high resolution, and rich way.

Guess I am trying to say that "good gamification" is not necessarily about being better at how we use rewards. If we are to take the TRUE deepest benefits of games, we would be better off looking outside of rewards completely. A reward system is nearly always an operant conditioning / external reinforcement approach. True, successful games are designed entirely around *intrinsic* motivation, with the most accurate definition of that phrase: the behavior is rewarding for its own sake, without needing ANY external reinforcement for the behavior. By definition, if it improves when rewarded, it was not intrinsically rewarding.

Self-Deterimation theory lives in a whole different world from operant conditioning. Gamification does not make this distinction, and does not seem to recognize the complex use of game mechanics in activities designed to be intrinsically rewarding, and thus mistakes the surface mechanics for the underlying core. Gamification is not necessarily wrong or bad because it is "shallow", it is wrong or bad because most of those trying to use it have misunderstood the science behind it. Not all "dopamine hits" are equal, just as the dopamine hit from a payout on a slot machine is an entirely different motivation from the dopamine hit one gets from designing, debugging, and compiling a wickedly hard, ultimately elegant code solution.
(sorry for writing such long comments)

Posted by guest on August 03, 2012 at 05:16 PM MDT #

I look forward to anything you have to say about this! You are such a breath of fresh air in this discussion. Thanks again.

Posted by Kathy Sierra on August 06, 2012 at 02:05 PM MDT #

Post a Comment:
  • HTML Syntax: NOT allowed

All things gamification, mostly focused on the Enterprise space but occasionally on other issues related to gamification. Thoughts are my own.


« July 2016