“Never, ever, think about something else when you should be thinking about the power of incentives.” -Charlie Munger
Most marketplaces I know end up seeing some unexpected behavior from their users. It’s always critical to be examining the incentives of all users of the platform. As soon as you have a clear understanding of the incentives, all of the behavior will make more sense. And believe me, as soon as you have any kind of scale, plenty of people will try to game your system.
Here’s are 9 common misalignments that I’ve seen in marketplaces. I’m sure there are many more.
1) The Perfect Feedback Disincentive – Let’s say I’m working on a marketplace and I have a perfect 5-star rating with a high number of reviews. I may be at the top of search results. At this point, I have negative incentive to get more feedback. I only risk harming my reputation on the platform so I’ll stop seeking any more feedback on the system.
The fix: Introduce some recency criteria (recent feedback is more valuable) and be careful of simple sorting on descending average feedback scores.
2) The Premium Placement Incentive – Premium placement (sponsored posting) has no place in marketplaces. Marketplaces should show the best available results to optimize for liquidity, not for paid placement. Paying for paid placement implies that you need a higher number of impressions to get a conversion. Good marketplaces are better off focusing on liquidity. And I know everyone will bring up Google and paid ads as an analogy. However, Google’s Adwords system is highly advanced and shows excellent results because they reward relevancy, most premium placement initiatives are simplistic fixed rates of billing.
The fix: Don’t offer premium placement.
3) The Bait and Switch Incentive – Many times sellers in a marketplace will list something for very low prices to ensure they show up in search results and attract customer interest. This leads to the inevitable bait and switch tactic.
The fix: Depends, but pay attention to search filters and choose wise defaults.
4) The Waiting Incentive – Take a look at a typical auction on eBay. It’s well-known that the price shoots up just before the bidding ends. So even with a 7-day auction, all of the action is in the last few minutes. If you see the item anytime prior to the last couple minutes, you have no incentive to bid. It’s wasted. You should just watch and wait.
The fix: Don’t allow behavior that obviously makes users want to wait to purchase.
5) The Membership Incentive – If a marketplace monetizes by getting paid memberships, then they typically restrict communication until they receive memberships. (See Care.com) However, this is another example of focusing on the wrong metric. If they were focused on liquidity they would want to encourage communication among users. Also, if someone is paying a membership, there’s a good chance they’ll want to make some good connections in the marketplace, disintermediate, and then cancel the membership.
The fix: Paid membership should be a plan B in my opinion. Better to capture part of the transaction.
6) Feedback Inflation Incentive – In some marketplaces where feedback is public, the incentive is to give very high feedback scores regardless of the actual experience. If my user profile is only associated with high feedback scores, it is easier to buy and sell on the platform. If I give negative feedback to someone, I may get retaliatory negative feedback. There is no upside for me to give a low score. (More here on how 5-star feedback systems are broken.)
The fix: Make some feedback anonymous and private.
7) The Spam Incentive – In a lot of marketplaces, quantity is rewarded more than quality. Craigslist is the classic example of this. There is so much spam on Craigslist because almost everything will get some impressions and clicks and activity, but there is no penalty to the users for spammy behavior.
Fix: there needs to be some downside to spam.
8) The Copycat Incentive – In public contest models, a la 99Designs, some strange incentives come into play. Basically, when your contest is public and other designers can see that you like another person’s design, everyone quickly copies that design to get a chance at winning the prize.
The fix: Don’t use a contest model, it causes a vicious cycle of quality.
9) The Desperate Incentive – This happens when a marketplace is so onerous in their application criteria that it’s very time-consuming to get into the marketplace. This can lead to an adverse selection problem since the best people don’t want to spend time unless they can see the potential reward.
The fix: Make sure your marketplace is designed to make it easy for the best people to start and excel.
Pay attention to these incentives and the behavior of your marketplace participants will be clear and you will be able to work around them gaming your system. Follow the advice of Charlie Munger and think of about the power of incentives.