From a grounded presentation by KPCB’s Vinod Khosla on what works and what doesn’t in venture capital, these two figures compare the pre-bubble and bubble funding model. One is fully funded before all the risks are out, while the other (pre-bubble) starts small and then sensibly ties new tranches of venture money to reduced risk. Guess which one works best?
[Update]Apparently some people are finding this chart hard to read, and not everyone is a Flickr member [ed. Can it be true?] to see the larger version linked to there. The gist: Pre-bubble it was common in technology to parcel out cash in increasing amounts at each stage as risk was reduced. So you might get a little at Idea, a little more at proving Technology Feasibility, more again at First Customer, and so on. In the late 1990s it became much more common to fund things all the way to the point at which expansion capital was required, thus meaning that the VC basically sat on his/her hands and prayed while time passed by.
The former financing technique is generally better (but not always), so Khosla’s point, and mine, in bringing it up, was that you could do worse that bring back stages tranches to financing tech. One poster here points out that fully-funded for some companies can come much sooner, with purchases happening early in the cycle. While that’s true, you can’t/shouldn’t build companies that way, planning, as it were, for a miracle exit. You’re always better to build the company as if you really mean it, not like you’re just messing with lottery tickets on an otherwise boring Monday morning.