A lot of companies ask me whether contests are a good way to spur social software adoption. In my experience, contests can be very effective in generating buzz, awareness, participation, and enthusiasm. They can also be demotivating and marginalizing. It all depends on how you run the contest, and the right way to do it is counterintuitive.

The theory that many people bring to contests is informed by classical economic theory: If you want to motivate people to change their behavior (e.g., to blog, create a wiki page, comment on a thread, etc.), it helps to give them an incentive. The bigger the reward (or potential reward), the more likely people are to participate.

In my experience, it doesn’t work that way.

I’ve seen companies run contests with relatively big prizes (e.g., a new iPod, a case of fine champagne), and I’ve seen companies run contests with small prizes (e.g., a $20 Starbucks card, a box of chocolates). In my experience, contests with small prizes are more successful than contests with big prizes. Small-prize contests generate greater participation, and that participation endures beyond the end of the contest. Large-prize contests generate a surge of participation during the contest itself, but that surge typically fades once the contest is over.

Why are small prizes are better than big ones? It’s actually not that uncommon. There’s a fair amount of academic research showing that rewards can actually negatively impact behavior. That’s because participants start responding to the reward rather than the other social, moral, or personal incentives they may have felt before the incentive was introduced.

In one famous experiment, researchers found that local residents were more inclined to accept a nuclear power plant in their town if there was no financial reward than if they were “bribed” to accept the plant. In another experiment involving negative rewards, researchers found that introducing late pick-up fines at day care centers increased the number of late pick-ups, as parents stopped viewing on-time pickup as a personal obligation and started seeing it as a financial trade-off. (Thanks to Barry Schwartz, who pointed me to the literature.)

These experiments support my own observation that social software contest are not successful when employees are motivated by a valuable prize. The wrong employees participate, their contributions aren’t very good, and they drop off as soon as the prize is gone.

I still think contests are a good way to stimulate participation. Contests can focus an organization’s attention on the rollout, and motivate new people to participate. But it’s critical that the prizes have value which is modest or merely symbolic. By offering a small prize, you can make the contest interesting and fun, without introducing the negative side-effects of a larger prize. You have to give them something to win–not because they want the prize for itself, but because they want to win. (I’m reminded of the movie Trading Places, in which multi-millionaires Randolph and Mortimer Duke destroy their nephew’s career and reputation over a $1 bet.)

So what prizes do I recommend? Here are a few ideas for modest gifts:

• A Starbuck’s or comparable gift card for $15-25
• A nice box of chocolates
• Lunch at a local restaurant
• A bottle of wine

If your company is a little more fun-loving, you can even offer something on the campy side, e.g.,

• A Neil Diamond CD
• A cheesy trophy of some kind
• A dozen cookies hand-baked by the head of the department

Above all, keep the mood fun and playful. Remember, you’re not appealing to people’s greed. You’re appealing to their creativity, their desire to have fun with their colleagues, and their drive for friendly competition.