Web start-ups almost always fail. So why are these guys launching one?

The story of America's greatest idea.
April 21 2010 7:05 AM

Risky Business

Web start-ups almost always fail. So why are these guys launching one?

Click here to launch the slideshow "Red Beacon"
John Dickerson John Dickerson

John Dickerson is Slate's chief political correspondent and author of On Her Trail. Read his series on the presidency and on risk.

Ethan Anderson, Yaron Binur, and Aaron Lee are in the failure business. Not explicitly—they don't run a pawnshop and they're not repo men. But they run a Silicon Valley start-up, Redbeacon, and as such they are statistically endangered. Depending on whose data you believe, the chance of their venture succeeding is between 10 percent and 30 percent.

For the privilege of waking up every day to those odds, each man has worked for a year and a half without salary, depleted his savings account by $50,000, quit a job at Google—the closest thing there is in this life to a Golden Ticket—and traded prosperous security for anxiety and the likelihood of failure. At Google, they enjoyed free gourmet food. At their headquarters, the only food I found on the day I arrived was a nearly empty box of Wheat Thins.

Advertisement

Redbeacon's office of 15 months is located down a dim hallway from a travel-book writer and Zena's Secret of Beauty spa in a San Mateo, Calif., strip mall. The suite consists of two offices and a conference room. There are more desks per office than a fire marshal would probably like. Phone lines cross the floor. Shirts from the dry cleaner hang on a lamp. To keep costs low, Ethan and Yaron and Aaron didn't buy professional whiteboards. Instead they picked up bathroom tile at Home Depot and drilled it into place with drywall screws. It feels like a hastily constructed bookmaking operation. If the cops got wise to them, they could wipe down the whiteboards and meet up in a cafe in Aruba.

Starting a company like Redbeacon is paradoxically risky. It exposes Anderson, Binur, and Lee to enormous financial risk, and yet the three are the opposite of swashbucklers. They are as cautious as they are frugal, rarely proceeding without typing out an agenda first. They spend hours scrutinizing even small product modifications on their vast flat-screen monitor. They embody the theory of the cautious entrepreneur outlined by Malcolm Gladwell in a recent New Yorker article. Redbeacon's founders took one gigantic risk to start the company, but they have spent every moment since attempting to minimize and manage risk.

Step 1: Find the Idea

The idea for Redbeacon was the result of a methodical search. Anderson and Binur had become friends at Google, where Anderson was the product manager in charge of launching Google Video outside the United States and Binur led the product development of Google News. Both were itching to start their own business. They got together twice a month to discuss what kinds of companies they could start. Each had to put together a presentation and pitch.

In June 2008, they finally found an idea that had enough potential to make them take the leap. At a Harvard Business School reunion a month earlier, Anderson had run into a friend who told him about a just-in-time labor service in Japan. Anyone who needed a part-time worker could contract immediately with available workers in the vicinity of the job. Anderson and Binur didn't think American labor laws would allow them to duplicate that, but they wondered if they could use the same concept to help consumers find local service providers faster and cheaper.

The co-founders of Redbeacon are buoyant and cheerful for people who work such long hours and face such long odds. All in their mid-30s, none of them cashed out with a big fortune from Google. They don't drive fancy cars. Binur lives with his girlfriend in student housing at Stanford, where she is a Ph.D. candidate. Money isn't nearly as important to them as the thrill of being under the steam of their own idea.

They talk about being entrepreneurs as a psychological trait, not something tied to a specific venture. "As an entrepreneur, you're always looking," says Anderson. Just a few years after graduating from Duke in 1998, he tried to start a service that that helped people who bought products from e-retailers return them quickly. It never launched, but the subsequent years in which he worked for other companies were merely a pause before his next venture.

While earning his masters at MIT, Binur helped launch a nonprofit that helps Israeli and Palestinian students learn computer science via joint classes in Israel. A citizen of Israel who spent part of his childhood in New York, he served as a staff sergeant in a paratroopers unit of the Israeli Defense Force. He traces his entrepreneurial roots to the immigrants of his home country. "Those who survived the Holocaust were the ones who left early and took quote-unquote risk. Those who stayed and wanted the good life and didn't want to take the risk actually died. Those who left took a lot of risk and came to this place that was a desert and built something. That really drives me."

Soon after Binur and Anderson conceived their plan, they brought in Lee, a Ph.D. in computer graphics, who was one of the founding engineers on Google Video. In August 2008 Anderson quit Google. The other two quit shortly thereafter. For the first three months they brainstormed before Lee sat down to start building the site and the analytic tools they use to measure its performance. Binur, who is also an engineer, lent a hand but his primary task is product management and strategy—understanding user behavior and luring service providers. Anderson took charge of sales and marketing.

To convey how quickly people would be able to summon professional help, they named their company Redbeacon, inspired by the police signal in Batman. When the authorities need Batman, they illuminate a beacon in the night sky. Just tell Redbeacon you need help, and assistance will arrive just as fast.

Here's how Redbeacon works: If I need a plumber Tuesday at 4 p.m., I type that request into Redbeacon.com, and shortly thereafter I get price quotes from plumbers who can make that time. I pick the one I like (based either on price or their quality ranking or a combination) and schedule the job. I haven't had to talk to anyone, and the job is at exactly the time I want. (No more waiting between 2 p.m. and 6 p.m.) Redbeacon intends to make its money by taking 10 percent from the plumber of every completed deal.

When the site launched its San Francisco service in October 2009 there there were only two part-time employees other than the co-founders. It offered more than 300 kinds of services, but the team chose to focus on 20 core services, including painting, photography, and cleaning. Starting from scratch, they signed up 1,000 service providers in a month. By December, they had several thousand, a sufficient pool of providers to run competitive auctions on each bid. Now, six months after launch, the challenge is to grow customer traffic. For competitive reasons, the company says they can't release their customer numbers to the public. But they will have to when they ask investors for money. The numbers are not likely to be huge, but the co-founders are betting they can make the case that with funding they quickly can be.

The plan is to expand to major cities like Los Angeles, New York, and Chicago. In their most ambitious moments, they aspire to replace the Yellow Pages, a more-than-100-year-old institution. As they considered the idea, Anderson, Binur, and Lee know they aren't alone in going after the $150 billion a year spent on local advertising nationwide. Their former employer, Google, is making a pretty good business out of that. Nor are they alone even in their small corner of the market. They have to compete against companies like ServiceMagic and Angie's List as well as start-ups like Thumbtack and HelpHive. But the trio was captivated by the scope of the challenge. They want to become a verb. "In the future I want people whenever they want something to simply say, 'Redbeacon it,' " says Anderson.

Step 2: Be Lean

During the Silicon Valley boom years of the late '90s, a company like Redbeacon would have put together a fancy business plan full of projections that were largely made up and sell their idea to a venture capital firm. They'd get a big investment and then they'd figure out how to build the company. The companies were "an ever shifting abstraction," as Michael Lewis wrote in The New, New Thing. Sometimes that abstraction fell apart before the huge launch party. Sometimes it fell apart afterward, when the business plan met reality.

Now everyone in Silicon Valley is much more cautious. The venture capital money has dried up. Big investments are rarer, and investors want to see proof of concept before they invest at all. Redbeacon is a lean start-up, a company that has launched small, in a limited way, and hopes to improve itself day after day, rather than in a single grand swoop. Lean start-ups try to build themselves the way hardware companies build hardware. Technology producers have long had a highly regimented system for building gadgets so that bugs can be discovered and fixed at each stage of development. Fatal flaws get caught before the big launch with the ice sculptures and B-list celebrities.

Redbeacon has also minimized its risk by not spending as much as it could have. It's much cheaper to launch a tech start-up in 2009 than it was in 1998. Redbeacon only employs part-time engineers and Web designers as needed because lots of software code is available for free. The only full-time employee other than the co-founders sells the site to service providers and handles customer support.

It also helps to start during a downturn. Anderson worked with Jobnob, a clearinghouse of engineers and marketers laid off from failed start-ups who are willing to work for free in exchange for future equity or a future job. The first time Redbeacon was mentioned in the newspaper was in a story marveling at how the recession had brought office rents to historic lows. Companies that started in the late '90s spent $20 million to do what Redbeacon has done with $100,000.

Gift Horse or Trojan Horse?

There may be a downside to the lean start-up. It may make new entrepreneurs too cautious. In Silicon Valley, people talk about the state of risk the way residents of a ski town might talk about the prospects for snow. If the new model of risk assessment makes everyone too conservative, the whole system might not thrive. Investors will only give a little money, and start-ups won't make grand plans. Adding an extra parachute before you jump may be prudent. But jumping out of the plane when it's still on the tarmac ruins the entire enterprise.

Not being risky enough is a challenge to Silicon Valley's economic longevity, and it's also a problem for Redbeacon. The company is threading a path between too much risk and too little. This was encapsulated by an experience they had in the early fall of 2009.

In the spring of 2009 they started considering whether to enter TechCrunch50, a sort of Sundance Film Festival for start-up companies. At this point, they had quit their jobs and opened an office but were weighing when and how to launch. As a condition of entering the competition, companies have to promise to launch at TechCrunch, which takes place in the fall, four months after an initial application process. In Internet time, four months is five years.

Delaying outreach to customers and the service providers was one risk. Another was that they could spend time competing for the prize and then fail in public.

About a thousand companies submitted applications, answering essay questions, providing a business plan, and taping videos. Then 50 companies were invited to compete for a $50,000 prize in front of a panel of technology journalists, venture capitalists, and CEOs. The prize money is nice, but the buzz that goes to the winner is why companies are willing to endure the long application process. If you win, you can get a big investment without much more than a good notion. The dream of every participant is to wind up like Mint.com, a TechCrunch winner two years ago that was just purchased for $170 million by Intuit.

Redbeacon was competing with companies like Anyclip, which allows users to search for any line of film dialogue and retrieves the exact piece of footage; CitySourced, which helps local governments use crowdsourcing technology to allow citizens to report potholes or graffiti; and Threadsy, which allows users to combine all of their e-mail and social networking accounts into one service.

The contest culminates with each company making an eight-minute pitch to attendees. That was Anderson's job, as head of marketing. He's had a lot of practice. He spends most of his day driving up and down Highway 101 to meetings, symposia, and networking cocktail parties. (He is like the bride at the wedding, always delighted to answer a question she's been asked a hundred times before: How do you get customers? Who are your service providers? What's your five-year plan? When are you going to start a family?)

Anderson had to explain Redbeacon in a way that would excite a very sophisticated audience. So he talked about cupcakes. He decided to order 500 for the crowd designed with the company logo. He called up the Redbeacon Web site, placed the order, and uploaded a digital image of the logo for the bakers to use as a pattern. He picked a baker out of the three who bid. As the final flourish, he put a call out to the cell phones of all Redbeacon users near the auditorium asking if any wanted to hire themselves out to deliver cupcakes. For $100 they found someone and the cupcakes were delivered to the crowd.

Obviously this entire process didn't actually take place in six minutes. (Anderson and his colleagues had picked the baker days before, and she spent the weekend baking.) But the presentation of how the site would work once operational dazzled the crowd. Redbeacon won and took home one of those huge checks like the ones Ed McMahon used to deliver.

As expected, the venture capitalists followed soon after. And here is where the story gets complicated. Books and blogs devoted to start-ups are filled with psychological counseling and advice about how to get in the door with VC's. Most companies dream of getting a VC meeting, and many never get one. In the month after winning, Redbeacon had 19 meetings with venture capital firms.

Investment from a VC would have launched Redbeacon. It could have moved into bigger offices—no more playing musical chairs to accommodate the sales team and freelance engineers. Anderson, Binur, and Lee could have paid themselves salaries and maybe taken a day of vacation. They could have advertised on city buses and hired an army of salespeople to sign up local plumbers, cleaners, and photographers. They could replace their saggy Craigslist couch.

But accepting venture capitalist money had its own risks. Silicon Valley veterans describe VC cash as "rocket fuel" or "a bomb." Anderson has his own metaphor: "It's like driving a car before the steering column is built and stepping all the way down on the gas. You go out of control and are destroyed."

Once the venture capitalists put in money, they expect the company to grow. They want to see the money spent on engineers, marketing, and expansion into other cities, all of which would presumably grow revenues. But what if the company isn't ready yet? What if you get huge and then you discover a fundamental flaw in your approach? You're too big to fix it in time. Your size and reach only means you've achieved the ability to broadcast your incompetence to a vast number of people.

Anderson, Binur, and Lee debated for three weeks about whether they should take any money. They interviewed other entrepreneurs and got mixed advice. Don't take the money so early, said many, you'll have to give up too much equity in the company in exchange. Launch small and then work out the bugs before you take on that big fuel.

Others argued they'd never have more buzz than after the TechCrunch50 win. The more buzz, the bigger the paycheck. The progress they could make during a soft launch would seem pathetic by comparison.

They decided not to pursue discussions that would lead to immediate money. This was their caution and their hubris. They weren't ready to handle the rocket fuel. Plus, they thought by launching on their own they could gather enough data about consumer behavior to make venture capital firms even more excited about what the company could do when it built to a larger scale. "We needed to prove to ourselves first that we could hit these goals," says Binur. "We need to prove that our models aren't theoretical. If we do that, we can go into VCs in a completely different position than after TechCrunch50. Then we had no data. We were just guessing."

A month after their TechCrunch victory, they launched in San Francisco and immediately started trying to understand how their customers behaved. Redbeacon's offices may be drab, but the software they've developed to analyze their data seems fit for one of those companies in all-glass buildings where everyone gets a Wii at the end-of-the year party.

Lee's software was designed to identify problems as soon as possible. "In order to be nimble and fast you need to have data on a daily basis if not sooner. A start-up's success or failure depends on how many iterations you can do within a fixed window."

What this means in plain English is that they need to identify problems quickly and adapt software to solve them. (At Google this concept was encapsulated in the phrase "launch early and launch often.") This is necessary to get the company on its feet, but it's also an important part of the pitch to future investors. New challenges don't stop coming once you've worked out the initial bugs. Investors want to see how quickly a company can identify and kill risks that might pop up in the future to threaten the money they might invest.

Manic data mining is also a necessity when you're operating on such a lean budget. With limited time, money, and attention, the Redbeacon co-founders can't spend resources guarding against every problem they might anticipate. Most of the problems they thought they'd face before launch didn't pop up. Those that did, they hadn't anticipated. Their constant risk assessment strategy means that as they look at every challenge they face they have to decide to spend their time only on those risks that threaten the enterprise. They don't worry about solving every edge case. "We don't build something until we feel the pain," says Lee.

The downside to moving forward so precisely is that the TechCrunch buzz is gone. Now they have to make the buzz all over again, with their numbers. How many San Franciscans are using their services? How often are they using them? How much are they spending? How satisfied are they? These data will make or break them. What if venture capital firms are underwhelmed by the data? The data might look impressive for the sandbox they've been working in, but there's no certainty the results can be duplicated when the company tries to get bigger.

The three men also know that you can iterate yourself to death. You can fix little neat problems and get lost in your technology until one day you wake up and someone has done your job better than you. "In Silicon Valley you know if you're not doing something, there are five other people in an apartment or garage who are doing it," says Anderson.

Since launch they have tried to build their customer base by expanding into social networks and partnered with local mothers clubs—working moms tend to book a lot of household services. They now think they are ready to take on the accelerant. At the end of April they will return to the venture capital firms and ask for money.

Nobody Is Going To Die

Anderson, Binur, and Lee have taken a huge professional risk, but it is not a life-threatening, or even career-threatening, one. The three co-founders could fail and step back into good jobs at Google or find some other company that would welcome them. Their experiences over the last year and a half only make them more marketable. Companies are always eager to hire entrepreneurs. "Although we lose our companies with alarming frequency," writes Geoffrey Moore in Crossing the Chasm, perhaps the most quoted book on starting a company, "we keep the people along with the ideas, and so the industry as a whole goes forward vibrantly, even as the names on our paychecks slide into another seamlessly."

None of the Redbeacon founders supports a family, and because only their own money is at stake, they would leave behind no angry investors if they failed. Since they have shown such prudence, future investors might look favorably on future ventures because they know their money won't be thrown after crazy risks.

Binur has an extra measure of perspective from his military career, in which he led reconnaissance missions into Lebanon. "I survived things that were much more risky," he says. "When I have to make a decision about the company going in this direction or that direction, yes, the company may fail, but nobody is going to die."

Fear of taking risks is influenced by two things: the gravity of potential failure and its probability. Though failure would not be that devastating for the co-founders of Redbeacon, they work as if it would be. They average 16-hour days and obsess about the placement of the smallest radio button on their site. That makes sense, given the way they view the probability half of the risk equation. "The statistics of being able to start a successful new company are so low, but when it's your own company you think, of course it will succeed," says Anderson. "You think secretly, these other people must have been idiots."

In some sense, Silicon Valley is the most perfect risk-taking environment ever devised. It demands performance and rewards the best insanely well, but it allows—even encourages—the losers to try again. Failure actually is an option.

Read the other profiles in this series, on  rock climbers Eli Simon and Pete Fasoldt, the band Girlyman,  and Marine Gen. James Mattis.

Become a fan of Slateand John Dickerson on Facebook. Follow us on Twitter.

  Slate Plus
Political Gabfest
Oct. 31 2014 1:29 PM You, The Gabfest, and a Hotel Room Win tickets to attend a taping of the Political Gabfest, live from David’s Chicago hotel room.