In most new markets, all you need to know is the number of zeros. Realistically, you cannot know more. Now, you can gauge the potential size of the addressable market, the number of customers that might be interested in the proposition, how much they might be willing to pay for something, and you might guess a rough market share that you can get. But there should be a wide range around each of those figures. What's important is not the ultimate value of the market, but is it big enough to be interesting?

I got an assignment in 1997 for an IT-services company to size the value of the Internet. We had no idea--we still don't know. But it was a fool's errand, because the answer was simply that it's big enough that you have to be there. Most new markets are like that. Now, there are a few exceptions: if you're launching satellites, or you're undertaking some major drug-discovery program, for example--then there a lot of upfront costs that you might have to commit to. But with most new markets you can stage your investments such that you can be pragmatic about investing a little, learning, and then doubling down.

The trigger for looking at new markets is looking at your own growth forecast. How much of that growth is realistically going to come from organic extensions on your current offerings versus tapping into new markets? For most public companies, a substantial part of the valuation comes from a new-markets phase that has yet to be defined. They have to go there to meet what the public markets are expecting in terms of their growth. And that will vary. But typically, if you're looking for upside surprise, it's going to come from new markets that people have not yet accounted for into your share price.