Paul Graham has a post out about how cheap its getting to start web companies and how this changes the landscape. Mr McClure over at 500 Hats sees this as The Destruction of Venture Capital. Don’t be fooled about what’s really going on here, folks!
As Mark Twain once said, “The reports of my death have been greatly exaggerated.” And so it is with Venture Capital: the reports of its death are also greatly exaggerated. Twain is also the fellow with the story about getting someone else to paint your fence, and that’s what’s really happening here.
Picture it from the VC’s standpoint. They need to make at least 10x to be happy because of all the risk in their portfolio. The 10x has to offset all the deals that went completely bust. VC’s understand (and many entrepreneurs don’t) that there is seldom middle ground. Your idea either flies or it crashes and burns and the investors lose their money. Valuation tends to undamped extremes. Your deal is either worth nothing, or it is being valued on an expectation that it will be worth vast sums at some point in the future. It’s all about perception, and that bubble is all too easily broken. The VC’s try to ride the horse as far as they can before the bubble bursts.
The astute observer will immediately start to wonder why the VC’s can’t make a buck by lowering their risk somewhat. If they can increase the likelihood of success for the deals they invest in, their returns should go up. But how to reduce the risk? Enter Darwinian selection. This is a subject I’ve talked about before. The idea is that when you’re dealing with unquantifiable risk, try a lot of different things with very little investment downside and then double down on what works.
In this case, VC’s benefit by having companies use either Angel funding (there’s now more Angel than pro money out there) or self-funding to let companies get started, get something in the market, and see what the appetite out there is for the idea. What happens next? The vast majority of companies get no further. They try out their idea and don’t get enough traction to matter. Some companies, like Don MacAskill’s SmugMug are successful enough that they never get around to needing the VC’s. What’s left are companies getting decent traction that need capital to expand. Suddenly the VC’s get to a richer feedstock of deals.
How should VC’s operate to take advantage of this world? Well, for starters, they need to be able to spread their Darwinian nets wide. That means looking at lots of deals. That’s why funds like Paul Graham’s Y Combinator are trying to be efficient at processing deals. They will want people to try to do as much as possible before coming to them. They will want Angels to fund as much as possible. Are you surprised then that VC’s like Graham are writing about this new model?
Is this really the future of VC: lots of little angel and self-funded deals that come for later stage expansion capital? No, it’s just one niche in their ecosystem, and it’s supported today because we’re in a period of experimentation as we try to identify the Web 2.0 big success stories. Very little software is needed to create such an experiment. There has been little time to start accumulating the long list of bells and whistles needed to be recognized as a player. The other thing that weighs on this is the relative paucity of liquidity for these kinds of companies. So many are being created as anyone who reads Mashable or TechCrunch will know, but so few are being picked up yet. If the period of experimentation comes to an end and there aren’t very many liquidity stories, the model will be in a lot less favor. As it stands, it is a relatively new model so everyone wants a piece of the new new thing. That’s normal with punctuated equilibriums.
I say it’s just one niche because lots of interesting software still can’t be created as cheaply as the next interesting web property. Consider Tim O’Reilly’s view that Web 2.0 is all about the intelligent back end. That again implies some significant technology (i.e. Google) that isn’t so cheap to create. It isn’t just a pretty face done in PHP very quickly. SaaS companies are also not built this way–it takes more capital. One reason is a difference of customers. SaaS is often B2B, and the business customer typically wants something a little more refined than a raw Web 2.0 startup produces. They’d like to hear that there are a few million dollars in the bank before jumping in. Businesses have a list of things IT has cooked up that you have to comply with before they’ll consider your solution, not the least of which is security. On the other hand, the transaction value is larger. Salesforce gets $50-$100 a seat month for a user. Google gets about $9 a seat month (their annual revenue / 12 / 120M unique visitors a month), which is considerably less, and they are the gold standard for these web properties. Yahoo gets less than half that.
Another example would be Zimbra, recently acquired by Yahoo. They raised a $5M Series A, $12M Series B, and a $14.5M Series C from big name VC’s. There are many other comparative examples.
Smart VC’s should be diversifying, that’s what asset management classes are all about, and it is an advantage that accrues to those with large funds. A certain number of deals in a fund should take advantage of the Web 2.0 + Angel Funded Darwinian selection. Another block should focus on companies building something more significant, perhaps for sale to business instead of broadly on the web. Those deals are often solving dusty old problems that are less exciting than finding the next potential Facebook in the portfolio, but there aren’t going to be so many Facebooks. One VC told me they were still adjusting to the idea that a Google-sized success comes along for the entire VC industry perhaps once a decade.
What about smart enterpreneurs? Choose your segment carefully. If you’ve got a great idea for the consumer-social web, it won’t cost you much to try it out. Don’t over engineer it. If you’ve got a business problem solved, be cognizant that the VC’s still want you to find some early milestones that require less capital to test the market fit. Also look for VC’s that understand the wildly different business models at work here in the final companies lest you be focused entirely on the wrong questions.
Related Articles:
Fred Wilson on Graham’s Article: He’s used the ugly “C” word: Commoditization.
O’Reilly says only a few Facebook apps get almost all the traffic. That’s what commoditization looks like. Beware total reliance on commodity web apps as the model of the future. The punctuated equilibrium will end, and just a few will be left as winners.