We invest in a lot of marketplaces and one of the questions we wrestle with a lot is determining the optimum commission level. Most of the time 10-20% feels like the right answer. I say ‘feels’ because it’s impossible to know for sure when you start out. At the 10-20% level you can make enough money to build a big business (particularly when non-transactional revenues are layered on top), but commission is low enough that it will be tough for competitors to undercut you. Bill Gurley makes this argument well in his seminal post A rake too far.
The chart above from Mary Meeker’s Internet Trends presentation published this week shows that 3-5% commissions have worked extremely well for a couple of Chinese marketplaces. As you can see they’ve enjoyed much faster growth than their US equivalents which charge much higher commissions.
We should be wary of assuming correlation implies causation, but there’s definitely something interesting to think about here. In the case where the market is huge and/or the underlying transaction costs are very low then taking sub 10% commissions merits thorough investigation. A radical way to look at this that might work in some markets would be using transactions as a loss leader to sell buyer or seller services.