The Next 10 Ideas for Billion-Dollar Software Companies

What if the world's best software entrepreneurs, the ones who could raise serious money, used all their special powers to take on big real-world projects, like schools, hospitals, labs and factories? This essay is based on my recent Tedx Windy City talk Take the Red Pill.

Since Netscape's public offering in 1995, more than 6,500 pure Internet companies have raised venture capital. But only 18 are worth more than $1 billion. Thirty years after the software revolution started, only 12 of the Fortune 500 are pure software companies. Twelve out of 500 isn't a revolution at all; it's the Massachusetts delegation to the House of Representatives.

It turns out to be very difficult to get customers to pay you hundreds of millions of dollars if you don't offer something substantial in return. Perhaps this is why our industry is undergoing an existential crisis, marked not just by the usual cyclical funding crunches but by a sense of weariness I've never seen in Silicon Valley before.

There are two venture-funded companies for shipping men’s underwear in the mail. There's an iPhone app for Apple fans to date one another, a polling service for Twitter followers to pick baby names.

Software matters more than that. As Netscape founder Marc Andreesen observed in his 2011 essay, "Why Software Is Eating the World," software is probably the most powerful lever for change the world has ever seen.

We just have to look at it the right way. The way Mr. Andreesen sees it, the world is becoming increasingly virtual:

Amazon rearranged its web site to promote its Kindle digital books over physical books for the first time. Now even the books themselves are software. Today's largest video service by number of subscribers is a software company: Netflix."

But wait, that's backwards. Yes the real world is in some cases becoming more virtual. But what's more important now is that the virtual world is becoming real. Amazon may be shipping electronic rather than physical books, but in my neighborhood, it's also delivering groceries via its own trucks. The company may lose the battle to deliver virtual goods, but it is winning the war to deliver actual goods.

Even the Netflix example cited by Mr. Andreesen isn't a simple story of the real world becoming virtual.

Consider the runaway hit "House of Cards," which Netflix produced after noticing that the customers who loved the 1990 BBC miniseries of the same name also loved Kevin Spacey. It wasn't such a stretch to imagine an American remake starring Kevin Spacey would touch a nerve with Netflix’s 30 million subscribers. Netflix used its virtual advantage of deep customer data to create a real-world advantage, made from physical sets and flesh-and-blood actors. In the viciously competitive streaming market, the real Kevin Spacey is what makes Netflix's virtual service stand apart.

What’s really interesting is what Netflix’s Chief Content Officer Ted Sarandos said about Netflix’s transformation into a live-action studio: “The goal is to become HBO faster than HBO can become us.”

In other words, the goal is for a virtual company to become a real company faster than the real company can become virtual. Linkedin is making the same effort with this essay, hiring editors and journalists to cultivate business writers: Linkedin is trying to become Bloomberg faster than Bloomberg can become Linkedin.

Who Will Win?

So what’s easier to do? Is it easier for a virtual company to get real or the other way around? Which companies will win the next 20 years, the new virtual ones or the old real ones?

To answer this question, consider another example. What was harder for Google: spending a decade to create an online map of every road and alley? Or strapping that map onto a car that drives itself?

It was of course, harder to build the map. Does this mean that 20 years from now, a company like Google might be a better auto manufacturer than a company like General Motors? Now that's crazy talk.

Except it isn't: Tesla, an electric car company now taking the world by storm, was started and run by computer scientists.

What other real-world incursions have been led by computer scientists? How about Nest, maker of a learning thermostat that starts heating your house half an hour before you arrive? It was built by the same folks who created the iPod. Five years ago, no one would have expected a high-tech entrepreneur to build a product for Home Depot's shelves.

An Airplane Manufacturer is a Software Company
Sure, but what about more traditional industries, like say, airplane manufacturing?

Well, it turns out that designing and building an airplane is mostly a gigantic software project. Each physical test of a wing costs more than $1 million. Supercomputers that were once limited to universities now model how a wing design will perform in freezing weather, after 10,000 flights, minimizing fuel consumption and maximizing safety. As a result, a million designs are tested online for every one that is actually built.

In other words, an airplane manufacturer is a software company, by a factor of a million to one. It is not inconceivable that a computer scientist, working with aerospace engineers, will one day lead such a company. Anyone familiar with Boeing's 787, now five years late and billions over budget, has to wonder when the Tesla or Nest of airplanes is coming.

Let’s get more down and dirty. What about drilling for oil? Surely this is an intrinsically physical process of resource extraction. No, it isn't. It turns out a technology called full wave-field inversion now allows oil & gas companies to find oil in places we had given up on or overlooked, which is the main reason why gas is back to three dollars a gallon. An oil & gas company is a software company.

Why Make Software for a Bank When You Can Just Be the Bank?

So does that mean that computer scientists can collaborate with folks in the real world to create new, disruptive companies? It does. Even the world of Mickey Mouse and Donald Duck will change because of these new partnerships. After all, why would entrepreneurs build software for the makers of animated movies, when we could be Pixar, making the animated movies ourselves?

You can ask this question of any industry. Why be in the dreary business of making software for banks when you can be the bank? Simple is a real-world bank co-founded by a refugee from the all-virtual world, Twitter architect Alex Payne, to use technology to make banking easier to understand and better for consumers.

Why, if you’re a group of Google engineers who figured out how to predict extreme weather, would you build software for insurance companies, when you can be the insurance company, directly insuring farmers yourself as Climate Corporation?

And of course an iPhone app like Taxi Magic existed for years, but no one cared about ordering a ride via the Internet until computer scientists launched their own car service, Uber, expanding one market at a time, taxi regulators be damned, and carefully controlling the entire customer experience from pick up to drop off. It turns out that the real world matters, and that technology for tracking driver locations and surveying customers about their experience can make the real world much better.

This has been the fundamental lesson of the company where I work, Redfin. Having built the first map-based real estate technology, our choice was whether to become a media company, selling ads to traditional real estate agents, or to become the real estate agent ourselves. I asked one of my only friends in Seattle at the time, Matt Bell, for advice.

I’d co-founded a software company before, so I expected him to tell me to run one now, sticking to an online-only business. But Matt, who played football in college, surprised me. He put a hand the size of my head on one of my little shoulders and said words I’ll never forget: “You were put on this earth to f--- with the order of things. Change the game. Change the whole game.”

The only way to change the reality of how people buy and sell homes is to operate in reality, not just virtually, hiring as employees our own real estate agents who are accountable to the customer, not the commission. We use technology to make the real-world process better at every step, getting people into homes faster and giving them better information on how to win the deal. We have saved consumers more than $100 million in fees.

What was hardest about this change wasn't hiring 500 real estate agents or taking customer calls at nine on a Sunday night. It was changing my conception of myself. When I went to parties, my old Silicon Valley friends would introduce me as a Realtor, not a software entrepreneur.

I learned from this that many of our so-called business decisions aren't made for business reasons at all but because of aesthetics: surrounded by computer scientists from Harvard and Stanford, we simply don't want to think of ourselves as booksellers, car dispatchers, or real estate brokers. This is the real reason software entrepreneurs for years shied away from starting real companies: it just wasn't cool enough.

Operating in The Real World Helps Us Build A Better Virtual World
But that's changing. What we're learning is that operating in the real world lets us build better software. For example at Redfin, our membership in local real estate boards lets our website show 20% more homes for sale, as much as a week faster. Our agents publish notes about the homes we see and the offers we write, sharing our real-world expertise in the virtual world.

And a better virtual world of course allows us to build a better real world. It's not unusual for our customers to cry when we complete the purchase or sale of their home. None of this would've happened if we'd sold ads to Century 21.

It almost didn't happen regardless. Though successful now, Redfin was turned down by more than 100 venture capitalists from 2006 - 2010. At the peak of the Facebook-crazed media boom, it became so difficult for us to raise money that we started each pitch by saying, “Here's what's weird about Redfin: we employ our own real estate agents. Hundreds of them.” Most meetings effectively ended right there.

What Scale Really Means
Why? Well for a long time, most venture capitalists wondered how a business that operated in the real world would scale. By scale, venture capitalists mean how quickly and easily the company's website can get traffic. Websites like Google and Facebook can get traffic quickly. To create web pages about say, TVs, media websites don’t have to actually build a warehouse full of TVs as Amazon does, or photograph them all, as Fab and Zulily would.

But what investors don’t mean by scale is what scale actually means: how large a business can be. Because it turns out that being able to deliver a TV to someone’s doorstep is a lot more valuable than letting someone talk about TV on Facebook.

Amazon generates double the revenue that Google does on a tenth of the traffic because Amazon is a software company that operates in the real world. The old joke about a million guys who walk in to a Silicon Valley bar and buy nothing -- the bar's declared a rousing success -- for years, this was how the most serious people in Silicon Valley evaluated a business.

From 2000 to 2010, would Amazon CEO Jeff Bezos have been able to walk into a venture capitalist’s office and ask for millions of dollars to build fulfillment centers across the U.S.? Or might he have been persuaded instead to build a social network for talking about, rather than actually delivering, books? Just such a company was started in 2006. Amazon bought it for pennies on the investors’ dollars.

The truth is that when real people are asked what kinds of problems they want computer scientists to solve, they almost always choose real-world problems. Nobody needs another website for sharing his status; people want a better wristwatch or a door they can unlock with their iPhone.

We know this because entrepreneurs can now directly ask regular folks for their support via websites like Kickstarter. Nearly all of Kickstarter's most popular projects are in the physical world, not the virtual world.

A New Fortune 500
But what if the world's best software entrepreneurs, the ones who could raise serious money, took on real-world projects bigger than a little gadget? Imagine how much better a hospital, a school, a research lab could be if doctors, teachers and scientists worked together with software engineers to create something not wholly virtual, and not completely analog either, but a combination of the two. The top companies in every industry, all 500 of the Fortune 500, are dreading the moment when a software-powered competitor digs into their world, into the real world.

This is surely what Mr. Andreesen really meant when he talked about what software would do to the world, and what the world would do to software. Software will eat the world, yes, but damn it, the world will eat software too.

(Photo credits: Ted Sarandon, k-ideas on Flickr; Google's self-driving car, ekai on Flickr; a Tesla, Randy Chiu on Flickr; an Airbus A380, KMJ Photography on Flickr)