A Startup's Second Act

American lives, F. Scott Fitzgerald said, have no second act. Most startups have that problem too. We bring our ideas to light, then fail, or get bought, or go public for a brief fizzy period.

But the greatest companies aren’t an idea, they’re an idea machine. These are the companies that aren’t just worth $1 billion, but $10 billion or $100 billion. Apple, the desktop computer innovator, now mostly makes gadgets. Netflix, the DVD-by-mail service, streams movies and creates its own shows. Amazon sells its e-commerce infrastructure as a cloud-based service.

Idea Machines

To be an idea machine, you have to invest in the machinery to deliver ideas. Bill Gates attributes Microsoft’s early success not to its operating system, but to the allocation of a third of its software engineers to build tools for the other two thirds. Some wondered back then how the company could set aside the time for that. But for the next twenty years, no one could make software faster. Now Facebook, which has all but built its own database software so that any engineer can immediately update the site without affecting the whole system, is in the same position.

An idea machine also invests in its idea-generating capacity. A great company’s deepest innovation isn’t in its product but in its being: a deeply idiosyncratic approach to hiring and working, accumulated through a million tiny acts of discipline and bizarre decisions, which alienates most outsiders, but draws a few in with the special, strange intensity of a children's story, as if it had been made for them alone.

You Have to Change to Stay the Same

Being that different is hard. The more a company succeeds, the more the consultants, bureaucrats and bandwagon-jumpers insist it needs their help to succeed: to grow up, to give up its youthful ways, to be reformed, to fit the business-as-usual conventions of other large companies. These folks want to keep innovation in the R&D lab, as if you could make different products without being different in other ways.

But the best companies become more, not less, idiosyncratic over time. They know that to do weird, they have to be weird. This idiosyncrasy has created a culture of beauty at Apple, of hacking at Facebook, of efficiency at Amazon. All of Apple’s ideas come from this culture of beauty. All of Facebook's from hacking. All of Amazon’s from efficiency.

Redfin, the company where I work, isn’t in the same league as these titans. But like many businesses that have had some success, we aspire to be. Our idiosyncrasy is in who we are, a software company that decided to hire a thousand real estate agents, working across the Internet and the real world to make buying and selling homes -- the search, the tours, the contracts -- better for regular people.

We’re now entering our second act. Our first was to make the existing process more consumer-friendly and efficient. There’s still a real estate agent who walks you through homes, but at Redfin she’s paid to put you first. Our technology for showing every home for sale on a map has now been widely imitated, but we also use it to make scheduling a tour or closing on a sale faster and easier.

The first act has generated hundreds of millions in revenue, and saved buyers and sellers hundreds of millions more in fees.

But to become a great company, we have to force ourselves to start this second act, which is a fundamental change in the process of buying and selling homes: how people see them online and in person, how they make offers, negotiate terms and close on sales. This act will take years to unfold.

In almost any play, the second act is slow. It starts out seemingly unrelated to the first, but actually the two are deeply connected: in our case, a hundred companies have tried and failed to change real estate without enough agents to cover the neighborhoods or without enough market-share to make up new rules. The only way Redfin has been able to earn the right to change the game has been first to learn the game.

In classical theater, as Fitzgerald knew better than anyone, that is what a second act is supposed to do: developing the idea from the first act into a much broader theme, which we sometimes don’t have the patience to understand or explore. Not having a second act isn’t about burning brightly and dying young, but taking the time while you’re still young to see what more you could be.

So just as things start to pick up at a startup, you have to slow them down again to make your story bigger. When the business becomes safe, you have to make it dangerous again. The truly great plays have five acts, with the best saved for last, and made possible by everything that came before.

Glenn Kelman is the CEO of Redfin, a next-generation, technology-powered real estate broker. Follow him on Twitter @glennkelman. Photo credit: Nitram242 on Flickr.