Marc Andreessen on his new fund, mega-franchise opportunities, and

Marc Andresseen, probably best known for starting Netscape and Opsware (now part of HP) and Ning. has just closed on raising $300mm for a new Venture Fund. He's on the board of Facebook and an angel investor in Twitter and is using the foundation of those experiences to look at 'franchise creating opportunities'. The following interview displays both is passion and irreverent sides. If you don't have time to read the whole interview from VCJnews, here's an annotated version of the discussion:


Q: What is a franchise company?

A: Andy Rachleff, who is a VC [formerly at Benchmark Capital] who now teaches venture capital at Stanford, did an analysis. Basically between roughly the mid-80s and the mid-2000s—a good cross section of time across a couple of different cycles—what he found is that basically there are about 15 companies a year that are founded in the tech industry that will eventually get to $100 million in annual revenue. Those companies in total represent a very large percentage of the returns to venture capital. His data show that they were 97% of all public returns, which is a good proxy for all returns. So those are the companies that matter. Those are the companies that have a big impact on the world. Those are the companies building foundational technology. Those are the companies that generate all the venture returns.

So, we’re going after those 15 companies and we’ll enter them at any point where we find them. Ideally, we’ll find them at the seed stage and we will help develop them and put in more and more capital as they grow. Failing that, if we screw that up, we will then, as we say, ‘correct our mistakes,’ and we would go into a later round.

Q: What’s your optimal deal size?

A: First of all, you’ll notice that there are two GPs. That’s a large ratio of dollars to GP, probably three or four times what you’ll find in a lot of funds. The deal-size range—this is an area where we’ve relaxed the focus—the deal size range is $50,000 to $50 million.

Here’s why. First of all, we think that it’s not a stage-specific exercise, at least the way that we do it. It’s a company-specific enterprise. We’re on the hunt for new franchise companies in the category I described. Our view is when you find one of those franchises—or a company that you think can be one of those franchises—you want to invest as much as you can both in time and effort across as many rounds as you can.

Q: How will that affect the number of investments you make?

A: The majority of the investments will be in seed stage. We may do as many as 60 to 80 seed stage investments in the first fund. Typically, in the seed stage, we wouldn’t go on the board. In fact, at the seed stage, we often advocate that the companies not even have boards. At the seed stage these days you’re talking about a company with like four or five people, so if you try to put a board together the board can end up with more people than the company. And a seed stage company’s mission in life, in our view, is to find product/market fit. Until it does that, all the rest of the company-building stuff doesn’t make any sense. That’s one side of things.

From our standpoint, a majority of our dollars will go into a much smaller number of deals at the venture stage than at the late stage. And so, hypothetically, 10 to 15 venture deals and two to three late stage deals or something like that. One twist is that we’re broadening the early to late side. We’re more multistage than most. Another twist is we’re narrowing the domain, which we discussed, pretty tightly.

Q: You seem to be a firm believer in the if-we-can-scale-this-thing, the money will come model. You’ve said as much about Facebook and Ning and Qik. But that doesn’t always work. It costs Pandora a lot of money every time it attracts a new user for example, but those users still prefer not to be fed ads. As an investor, how do you know when that’s a viable approach?

A: No. 1, the details really, really matter. The cost structure really matters. I don’t know enough about Pandora. When people get in trouble with this sort of thing, it’s usually for one of two reasons. Either the market wasn’t going to be that large—in which case deferring revenue to get to the market wasn’t worthwhile—or the costs are just too high. So, I don’t want to talk about Pandora specifically.

But, in contrast, here’s how I think about the economics of Twitter, for example. The economics of Twitter are that they’ve spent about $15 million. They have created already a global brand name. Ben likes to point out that the Bing ad campaign [by Microsoft] is $300 million of advertising. Would you rather own the Bing brand or the Twitter brand? So, what’s that worth? Two, they have a user base of about 30 million users now, growing very fast. And, three, they have that growth rate, so they have all the future acquisition.

But even just looking at the current user base, they’ve spent maybe 50 cents per acquired user. Total. For everything. All development, all marketing, everything. And so on the revenue side, you say: ‘Suppose they want to monetize that? Can they get 50 cents per user per year in terms of ads?’ Yeah, probably they could do that. So to go get that $15 million back seems really easy, and it seems like there’s a lot more upside beyond that.

Finally, here's a link to his blog (now on hiatus).