I am interested in marketplace businesses. I was the first employee of a company called SharkTank (not the TV show; an online labor market for legal services) in 2000. Since then, I’ve spent most of my time as a venture capitalist, having invested in marketplace businesses DealerTrack (auto loans), Solvate (labor, again), ADstruc (out-of-home advertising), LendKey (consumer loans), and Worktopia (meeting space). Some of these worked out, some didn’t, and some are current, active investments.
I find myself repeating some of the same lessons from these experiences in conversations with entrepreneurs starting marketplace businesses, so I thought I’d capture some of those thoughts here. Note much has already been written on this topic, from blogs (here’s a great post by Bill Gurley of Benchmark) to HBR articles. If you are interested in these types of businesses, I encourage you to read more.
I like marketplaces because they are what I call flywheel businesses. Once they get going, they have momentum all their own and are hard to stop. The unfortunate corollary, though, is that the same inertia that can make them juggernauts makes them that much harder to start than many other businesses. The core problem at the outset is whether and how to attract buyers or sellers to the platform and, given that it’s usually impossible to instantly attract a large volume of both, which to pursue first. This is often called a “cold start” problem.
Here are a few strategies that can help you overcome this chicken-and-egg, or cold start, problem:
- Play matchmaker. Marketplaces are prone to “build it and they will come” expectations. Technologists, especially, often focus on building a great product – a very efficient marketplace – and expect that the first buyers and sellers will happily use their new, improved, transaction automation engine. The reality is that, many times, management teams must act as matchmakers to facilitate the first transactions. If the buyers and sellers are businesses, you may need to proactively call up the first would-be buyers once you convince a seller to try out the platform. For consumers, you may need to do some handholding over e-mail and/or chat for the first transactions, in addition to being very smart about your product instrumentation, marketing and product iteration efforts.
- Focus. Initially, target your efforts at a particular type of buyer, seller, and transaction. eBay started as a marketplace for certain kinds of collectibles.
- Gather inventory from everywhere. You can’t facilitate transactions until you have inventory. At the outset, though, don’t wait for sellers to show up one-at-a-time to provide it. Scrape it from third party websites (if you can do so without violating their terms of service). Pay for partner inventory. Use interns or other sources of cheap labor to search the web or hit the phones. In short, get creative.
- Under promise to early adopters. Many marketplaces get started by encouraging sellers to list inventory and wait, arguing that buyers will eventually show up and there’s no harm in waiting (so long as the inventory is not perishable and the sellers have other channels to market). This pitch can work, but can backfire if the buyers don’t show up in the timeframe that the sellers are told to expect. Don’t make promises you can’t keep, and meet any expectations you set.
- Provide another product. This strategy is last on my list because it conflicts with the more-important “focus” manifesto and is extremely hard to pull off. In some cases, though, for marketplaces that can take a particularly long time to develop, it may be necessary to offer other product(s) to buyers and/or sellers. These products should provide value to any one user, whether or not transactions are flowing through the marketplace. ADstruc, for example, provides a variety of productivity, workflow, and data products to its customers. When these products encourage the addition of inventory to the market and/or facilitate transactions, they can be successful.
None of these strategies are foolproof, but nor are they mutually exclusive. On the contrary, getting the flywheel going will probably require the implementation of a few of these approaches for many marketplaces. By the way, here’s another great blog post on this topic by David Beisel of NextView Ventures which echoes some of the same themes; I wish I would have found it before writing this one!
