Vision Plus Experiments
Khosla: Be obstinate about your vision, but be really flexible about your tactics. What that means is — it's important to have a vision of where you're trying to go. How you get there is a series of experiments. You can call them failures when they don't work, but they're learning experiments where you try things, see what works, try the next thing.
Entrepreneurs don't try to get there in one leap. They see where they are, what resources they have — it may be nothing except their college money. You take a step in the right direction so you can garner more resources. Try the next experiment. It may fail or succeed. You try something new, garner resources for the next step.
How you get there is a series of experiments. You don't do the whole thing in one step — that's very high risk.
Too Much Money, Too Much Confidence
Khosla: People who raise too much money get too much confidence in their own plan, which is normally BS. The right way to do a business plan is flex planning — have a vision, but plan the next 3 to 6 months. Test every assumption you have. Even if you're sure it's true, test it if you can in the marketplace, in running an experiment.
A business plan is about testing every assumption.
Too much money gives too much confidence in a plan that's normally BS. Test every assumption.
Base Camp on the Path to Everest
Khosla: You want to shoot for Mount Everest, but nobody ever got to Everest without getting to base camp first. You build a stable place where you've learned a lot, where you can tread water for a while, rest up, build a business. Maybe you get to cash flow break-even at base camp.
But you want to set up base camp where it helps you scope out the path to Everest. That's the whole point. My problem with VCs — in the process of trying to get a cash-flow-neutral company, they don't worry about that base camp being on the path to Everest. They'll set it up anywhere, and then it doesn't help you get to the big vision.
Build base camp where it helps you scope the path to Everest. VCs set it up anywhere — then it doesn't help.