A key component of venture capital is diversification.

Some people have made investments in startups, and they’ve taken off for a while, and then failed. If you invest in one or two companies, you have a high risk that you are going to lose your money.

If you look at startups on a broad basis, what’s the chance and probability that they will succeed? What’s the chance and probability they will fail? What you find is about one in 10 will outperform all the other investments. If you invest in 10 companies, one of those companies will return such a high return that you have positive return for your portfolio as a whole.

The concept of diversification basically says: “Paying attention to that statistic, if you invest in only two or three companies you’re not going to get the return you want.” The chances that you’re going to find that one that is successful out of the group is low.

We find that it makes sense to invest in 20 or more companies that have all been thoroughly diligenced. If you invest in 20 or more companies it provides you enough winners and losers, such that you had your risk, so that the few that do really, really, really well overcome the many that will likely fail.

Diversification is a critical component in the logic of venture capital. I used to manage the R&D process at Boeing and we invested about $300 million in many discreet projects over the years. This was for technology that was to be used in our product line five years out, 10 years out, 15 years out. We had the best PhDs, the best scientists, in our industry, and you would think we must know exactly what the best choice is. Running that process for many years, I can say that “no,” we don’t. We can understand the science, we can understand the physics, we can understand all the mechanics of business. That part is invention.

Innovation is the customer seeing the product, and the product working the way the customer wants, and the customer being willing to buy it, again, and again, and again.

The challenge investors have in understanding exactly how the market will fit with the product, is really quite a big challenge. The product of diversification still says two ideas out of 20 are going to hit it. Two out of 20 times the customers are going to love it, and it is going to take off.

The principles that we apply are, one, make sure you have the right scientists, the right business people, and look at the transaction to make sure that on paper it looks good.

Then, two, invest in 20 or more companies so that you get enough shots on goal with customers to finally figure out what the customer product mix is. It is a challenge, and it is frankly hubris to say as an early-stage company that you know better than the market which one is going to win.

I think even the best venture capitalists will say that the things that have been the best performers in their portfolio were not the companies they thought would be the best performers. And some of the worst performers in their portfolio, they thought were going to do great. But they invested in enough companies with enough diversification so as to achieve a good return for their investors either way.