During dark times, real entrepreneurs come out. They are not competing with 10 look-alike companies for engineering talent, so it's a great time to invest and help build companies.
If you choose a market that already exists, say, networking equipment, you have to compete with an established company like Cisco. Even if your product is marginally better, Cisco can fudge it and outsell you.
I'll say this: I can't think of one instance in my 20 years in venture capital in which I have wanted to sell a company before the entrepreneur.
We work as a team. I think having the individual being shown as a star actually creates problems internally. We encourage all our investors to work as a team for the benefit of the founders.
In a globalized world, one application can spread like wildfire and there's only one winning company, which means you have to invest more than you've ever had.
The trick is, a market has to be nonexistent when you start. If the market is large early on, you will have too many competitors. You have to make it large.
We have co-opted seed funds. You know, Y Combinator, that was completely our money. We have secret handshakes with a whole bunch of people. Very dangerous, because word gets out that so-and-so's money is Sequoia's money, that would not be a good thing.
Raise as little as you can to get you to something that you can show - plus maybe a quarter or two so you have a little bit of cushion - and then raise some more money. Raise as little - not as much - as you can because that's the most expensive equity you're going to sell.
You meet with a CEO or founder. You talk about sales, engineering, product management and give some ideas or suggestions. And the founder quickly understands that you really can help them both operationally and from a strategic standpoint.