Paul Graham is at it again with one of his lucid and thought provoking essays, "The Venture Captial Squeeze", on why he thinks venture capitalists are hosed. His argument basically that capital is less important due to a few factors which make things cheaper to start companies and that Sarbanes-Oxley compliance makes it harder for startups to exit through the public markets (which implies they must sell for less to bigger companies in order to ensure a VC return).
I don't think that anyone can dispute the fact that post-bubble we are in a much different IPO market and I would not argue with the fact that it is cheaper to start companies these days but Graham misses a couple of important things that bear mentioning [Full disclosure: I have tried both the bootstrapped approach in the past (didn't work) and am currently at the beginning of a VC-funded go at-bat].
First, he forgets to mention that even having fewer people can get expensive quickly. Unless you are going to fund college students that can live in dorms and eat Ramen for dinner (something that I think he is having some success at, most people have real expenses and thus can't afford to work for no pay, or even minimum wage. If you factor in basic benefits and other people-related overhead, it is not out of the norm to think that you'll spend about $700K for a small team to come together to develop something worthwhile over the course of a year (see "runway" below before you quibble with this timeframe). And that amount is just above the reach of most angel or friends-and-family investors, or at least it is if you want to keep the signal-to-noise ratio of dealing with investors manageable (someone once told me that having angels do more than $200K was like trying to herd a litter of kittens due to their number and how fragile they can be).
Second, and this is something that he overlooks altogether (as do all of the drum-pounding Web2.0ers), when you are a bootstrapped startup that fails you can run the risk of being the tree in the woods that no one hears fall. When we did Memora we got a bunch of press at launch, great product reviews, and loads of interest. And yet, after we failed the only thing that people wanted to know when I would talk about it is: "who funded you?" as if this was more important than anything we actually built. This is lame I realize, but a fact of life you need to consider when you just look at the low odds of success in the average startup. Here the "venture filter" is your friend for two reasons: first it means you've made it past a set of guys who specialize is weeding out potential failures (which more than picking successes is what is at the heart of a top-tier VC fund), and second it means that if timing, the market, or a hurricane causes you to fail, you won't be that tree that no one hears falling.
Overall though, I think that Paul is right on to point to the fact that the VC model needs to change, at least in the web space. I think the best guys get this and like to think that our current venture is an experiment in this new direction.
The important thing that good VCs already realize is that the way to deliver value to entrepreneurs is shifting. Entrepreneurs no longer need access to big amounts of capital to grow; we don't even need access to expertise about how to grow or who to hire (because teams are smaller and the Internet is a wonderful way to identify, vet, and attract the right set of folks).
What we need is runway and support.
The first comes directly from a judicious amount of capital. Assume for a moment most startups do not get their first product right. Having enough capital to refine the product, the messaging, and even the team composition if necessary really helps to overcome the mosquito effect that affects most bootstrapped startups (you get one sting and then blam, you're dead). Ergo the number cited above for keeping a team running for a year in tight but manageable conditions.
The second comes from rolling up the sleeves and getting down and dirty. Now, every VC will tell you that this is something that comes with their money but it is just not true. First off, most VCs don't come from operating backgrounds which means that when it comes to helping you position your product and important "business" things like that most of what they can provide can be gleaned from the pages of The Portable MBA. This doesn't mean that VCs can't be valuable thought-partners though-- after all a VC that has been in the business for a couple of decades has seen plenty of good operating guys go through several technology phases.
The real killer is that most VCs spend the largest share of their time either hunting down new deals (something that becomes more important when you've got a billion dollar fund and are making ever smaller investments) or dealing with their existing companies that are on fire. Therefore time for thinking and really engaging at the point where they can be most valuable to small companies falls by the wayside. And unfortunately the dynamic that underlies Paul's essay (falling costs of starting companies) is going to make this worse before it makes it better.
So the big take-away is this: if you're still thinking about starting something and think that the reasons I've laid out above make it worth your while to at least explore the VC route, make sure that you pay close attention to whether you'll be getting the runway and support you'll need to get it right. And, of course, go out and read Graham's essay because it's written with the little guys in mind.