Marc Andreessen Answers the Tech Valuation Question

The venture capitalist says many investors don't grasp what's changed.

Remember Netscape?

Photographer: Scott Ellis/Bloomberg

Are you a venture capitalist, angel investor or startup entrepreneur? Do you seed new technology firms, or write about them, or work for one them? Then you are probably spending way too much time and energy obsessing about valuations.

So says the man who invented the first commercial internet browser, sits on the board of Facebook Inc., and helped back companies such as Airbnb Inc., Box Inc., Groupon Inc., Instagram Inc., Lyft Inc., Pinterest Inc., Skype Inc. and Slack Technologies Inc.

I sat down last week with Marc Andreessen of venture-capital firm Andreessen Horowitz to hear his thoughts on unicorns, technology, entrepreneurship, psychological scarring of investors, the future of e-commerce, Arpanet and really, just about everything else interesting under the sun. His insights are as profound as they are counterintuitive.

He thinks of himself as an engineer first, a problem-solver committed to “figuring out how things work -- and then figuring out how to make them better.” Some of the key points he made follow:

The Math of Venture-Capital Returns

Stock returns tend to be driven by a handful of big winners; for venture investors, it’s even more lopsided. Venture-capital funds typically have a 50 percent failure rate -- half of the investments lose money, with half of those being total losses. The third quartile breaks even, or returns two or three times their money over five to 10 years. The real action is in their top quartile, which can generate return on investments of anywhere from three- to 1,000-fold.

Andreessen observes “We make our money on the ones that work and our reputations on the ones that don’t.”

As an example, consider a fund that owned both Segway Inc. and Google Inc. The self-balancing two-wheeled scooter was an expensive flop. The fund lost one times its investment in Segway; investors in the same fund saw their investment into Google generate gains of roughly 10,000-fold. With returns like that, venture investors don’t mind kissing a lot of frogs looking for their prince.

The top-performing venture capitalists embrace the “Babe Ruth Effect” -- they have a lot more strike-outs, but also a lot more towering home runs.

Stop Worrying About Unicorn Valuations

The obsession with a handful of unicorns, or startup companies valued at more than $1 billion, reflects the tail wagging the dog, and not a deep understanding of the impact and scale of new technologies.

Media headlines have been shouting that tech is overvalued every year since 2004; many observers have been insisting a tech crash is imminent for more than a decade.

Andreessen disagrees: “There are no bad ideas, only early ones,” he says. Take the very worst concept you can think of from the dot-com era -- say,, the poster child for the excesses of that time. Andreessen brings up online pet-food retailer LLC, the most obvious successor; it was recently acquired by PetSmart Inc. for $3.35 billion in one of the largest e-commerce acquisitions ever

Tech Infrastructure Has Changed

During the 1990s, startups faced a pretty huge headwind: Setting up and running a website-based company could cost as much as $10 million. It required a full technology team to buy and maintain servers, databases, routers and communications infrastructure. Today, this entire system can be purchased for a tiny fraction of that from cloud providers such as Amazon Web Services.

Lowering this huge barrier to new technology ideas has had an enormous impact; start-up costs have been cut by a factor of 1,000, according to Andreessen. For many new tech businesses, the initial investment is nothing more than the founders’ laptop computers and the cost of an internet connection.

This makes it possible for venture capitalists like Andreessen to engage in a “Darwinian process of running experiments, exploring an immense number of ideas,” he says.

Software Is Eating the World

In 2011, Andreessen had an op-ed in the Wall Street Journal outlining the centrality of software -- not hardware -- in the tech sphere. It was more prophetic than most people could have imagined. Consider Apple Inc.’s iPhone. The Chinese contract manufacturers that build and assemble the phones make a few dollars on each item; Apple keeps the vast majority of profits. Indeed, its software is so valuable that the company has captured almost all of the profits in the mobile-phone industry.

A parallel case exists for electric carmaker Tesla Inc.: It doesn’t yet manufacture the batteries. The genius, though, is in the software managing how the power is extracted from each car’s lithium-ion cells. That is before we even get to the company’s nascent autonomous driving and self-updating operating systems.

But what’s the right comparison? Is Tesla today like Apple in the mid-1980s, when it introduced the first popular desktop with a graphic user interface that was soon copied by everyone else? Or is it more akin to Apple circa 2007, when it introduced the iPhone -- a device that competitors have failed to match for overall user appeal, helping to make the company the world's most valuable business? The answer to that tells you whether Tesla is under- or over-valued.

Regardless, the point is that software, not hardware, is paramount.

I spoke with Andreesen for 90 minutes, and the full interview is featured on this week’s Masters in Business podcast. You can stream or download the full show, including the podcast extras, on iTunesSoundCloudOvercast and Bloomberg.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

    To contact the author of this story:
    Barry Ritholtz at

    To contact the editor responsible for this story:
    James Greiff at

    Before it's here, it's on the Bloomberg Terminal.