I'm a big fan of marketplace businesses (I've invested in Red Beacon, Rubicon Project, etc) - it's easy to theorize these as big opportunities with strong competitive barriers via the network effect but such businesses are actually very hard to make work. HBS Professor Andrei Hagiu has studied online marketplaces (which he also calls Multi-Sided Platforms - MSPs) and has some good thoughts on the nuances and difficulties. He shares his thoughts below - a must read for entrepreneurs contemplating a marketplace business:
Multi-sided platforms (MSPs) tend to be some of the most attractive businesses to build and invest in. These are platforms that create value by connecting multiple types of customers, which results in network effects (e.g. eBay, FPlayStation, Airbnb, oDesk, Red Beacon). I’ve written extensively on how to design MSP strategy, expand them, etc. But my impression is that this is yesterday’s news. By now most people understand that MSPs can be very powerful and everyone wants to build one. We’ve gotten to a point where the MSP-mania is creating a lot of misguided entrepreneurial attempts or investments:
- Superficial claims of MSPs and network effects where there are none (e.g. online brokerage start-ups during the first Internet bubble)
- Even if there are network effects, they are not supported by high switching or “homing” costs (e.g. Groupon)
- The relevant market is lacking some key institutional “infrastructure”, which makes MSPs the wrong model to go after. Example: starting in 2005 a lot of entrepreneurs and investors were bullish about market exchanges for IP (e.g. Tynax, Yet2.com, Ocean Tomo auctions, The Dean’s List). In principle, creating MSPs for a highly illiquid market makes sense. But I always thought it would be impossible to get MSPs for IP to scale given the lack of valuation transparency and lack of transaction standards (not to mention huge litigation risks). Today, all of these attempts to create eBays or Craigslists or Yelps or Sotheby’s for IP have either been shut down or are functioning at puny scales.
- Even if there is a sound potential for MSP with network effects, it’s gotten much harder to build them. The biggest obstacle today is not the chicken-and-egg problem, but resistance by other players in the relevant industry. The ability of MSPs to extract value from everyone else in the ecosystem is well-known and nobody wants to end up a tiny asteroid orbiting a big planet (like PC makers around Windows or music studios around Apple’s iPod). Great example: Brightcove. When I first talked to J. Allaire for my 2006 case study, the plan was to become a 4-sided platform for Internet video. BC would intermediate transactions between video content providers, users, web affiliates and advertisers. By 2009, the company was forced to scale down its ambitions: big content providers didn’t want it to become the equivalent of iTunes for Internet TV. It’s still successful as an “enterprise software” provider to content providers, but its valuation ambitions went down from $1.5 billion to just several hundred million…
Several important implications for building/investing in MSPs today:
- Try to find spaces in which alternative intermediation models to MSPs might be more valuable and easier to build. Those are more suitable opportunities for start-ups in particular since less threatening to large players. Examples:
- Brightcove in its current incarnation: not MSP but “one-sided platform”
- Gazelle and EcoSquid in re-commerce: taking advantage of eBay and behaving as “merchants”, not MSPs. I’m currently doing some work with Gazelle’s CEO: becoming an MSP is very tempting but would be the wrong path to take at this point.
- RPX in IP: it buys IP and then licenses it to clients (merchant not MSP). Much more successful than the alternative MSP models because it avoided the chicken-and-egg problem, does not depend on external IP valuation mechanisms and does not threaten large tech companies. RPX was founded in 2008, went public in May 2011 and is now worth over $1 billion. The other related example is of course Intellectual Ventures (I have a case study on them too).
- Even if the ultimate goal is still MSP, one needs to be smarter about how to get there:
- start off as a one-sided platform serving one customer group and then realize MSP potential
- start off as merchant then become MSP
- Network effects and MSPs just based on business model are no longer enough. To build sustainable businesses and barriers to entry, one needs to also solve hard technology problems (either software, hardware or strong IP). Too many MBAs shy away and try to compete just with “clever” business models.