A Startup to Help Build Startups
Photography Courtesy Upstart
It’s an old story. College seniors head toward graduation, many with dreams to start their own businesses. Then reality hits, and they end up taking jobs in fields that promise a steady, low-risk paycheck. The startup option goes up in smoke.
That scenario bugged Dave Girouard. A little over a year ago, Girouard, who was then president of the enterprise unit at Google (GOOG), was talking with friends and family who saw that drama play out again and again. Grads with computer science degrees from top schools and a flair for Ruby programming have no trouble finding work they love. But for smart, creative seniors who don’t have those specializations and who carry big school debt and have risk-averse parents, any idea for a startup dies in its cradle.
Girouard had some personal experience here. It was a stretch for his family when he got into Dartmouth. “We could in no way afford it, but my father was adamant that I should do it no matter what,” he says. He graduated with the requisite “ton of debt,” which only rose when he went to grad school. When he got out, he took the most lucrative job he could find at Booz Allen Hamilton (BAH). As soon as he paid what he owed, he quit to join Apple (AAPL).
Debt and the decisions it drives are a problem for the students—obviously; who doesn’t want to follow his bliss? But it’s also an issue for society at large, which misses out on the job creation that can come from small businesses. It bothered Girouard that investment banks and law firms became the default destinations for grads who would have liked to strike out on their own. So he started talking to experts—economists, academics. He looked at government statistics. He looked at crowdfunding sources such as Kickstarter and Indiegogo. He decided he wanted to build a startup to help startups, so he resigned from Google in March and launched Upstart in April.
One of his findings is that it doesn’t necessarily take a fortune to help get someone started. Many of these kids couldn’t, without credit cards, put their hands on $30,000 if their lives depended on it, but if they could, it would be enough for them to rent office space, buy computers, hire some help, Girouard told me last month. And that at least gives them a shot.
The Redwood City (Calif.) company, now with eight employees—helps to match promising upperclassmen from select schools—including Rhode Island School of Design, Dartmouth, University of Michigan—with patrons who have money to invest and business experience to share. (Upstart just added Berkeley, Harvard, NYU, Stanford, and Yale to its roster.) Thirteen mentors signed up initially, including Frank Moss, professor at the MIT Media Lab and a longtime industry exec; and Andy Palmer, who co-founded Vertica Systems (HPQ) and VoltDB.
It should be noted that Girouard himself left a solid gig at Google—something his financial adviser counseled strongly against—to launch this effort with backing from Google Ventures, Kleiner Perkins Caufield & Byers, NEA, and Mark Cuban. Girouard was joined by Damon Whitsitt, another former Googler, and a few others. They wanted to find a way that investors would share in the entrepreneur’s upside but that would also limit the entrepreneur’s downside risk. If the startup doesn’t work out, the investors don’t recoup their investment, but the feeling is they have enough money and expertise to sustain the loss.
Again, the idea was to fund the person, not the startup, paying X amount for a certain percentage of that person’s income—regardless of its source—over 10 years.
In coming up with a plan, Girouard initially thought of some sort of auction: “Here’s Bill, who’ll graduate from Stanford with a computer science degree. What’s 3 percent of his income worth to you?” and have the investors bid it out. One problem was figuring out how much a given person’s earning potential would be over the period in question.
Right around then, Girouard heard about “a kid” in New York who was working on a way to forecast an individual’s earning potential over time. This was Paul Gu, one of the first class of Thiel Fellows. He had been a double computer science/economics major at Yale but left after three years to pursue his research. By the time Girouard reached him by phone, Gu had aggregated a couple of publicly available data sets to help in this task.
Why was Gu pursuing this? Because his friends were taking jobs they didn’t want because of student debt, he told Girouard. Bingo.
Within days, Gu flew out to meet with Girouard and Whitsitt. “As soon as I saw what he was doing, I thought, why the hell do an auction? Let’s just price this ourselves, and we can build a model without burdening the backers,” said Girouard.
Gu came aboard as Upstart’s first non-Google employee. From that time on, Gu and team have been gathering data from the government, from schools, and from private sources that collect salary data. They recruited prospective grads and backers in parallel. Since the pilot program went public in August, nearly 300 backers/mentors have signed on, Girouard said, and Upstart is launching road shows at college campuses this fall to talk up the program.
Basically, the applicant asks for a set amount (funding is capped at $50,000). Upstart then uses its “proprietary models and analytics to figure out the fraction of their future income they must agree to share to raise their desired amount,” according to its website. The recipient gets X dollars for each 1 percent of income shared up to a maximum of 7 percent. No payments are expected when the recipient makes less than $30,000 per year.
As an aside, Upstart structured an agreement that creates incentives for Gu to return to Yale to finish his degree—which Gu said he intends to do.
This is not a totally philanthropic effort. Upstart gets 3 percent of the patron’s initial contribution. And should the effort pan out, when the entrepreneur pays back the patron, Upstart collects a 1.5 percent service fee.
Both Gu and Girouard brim with enthusiasm about the effort. They feel they are doing well by doing good. Coming up with the analysis is challenging. And they all seem to like each other. “I like to call it the stuck-in-the-airport-together-for-four-hours test,” says Gu. “We’re happy to spend time together. A lot of time in the Valley, there’s this culture of ‘we’re so super smart that we know everything.’ Well, we definitely don’t have that here. There’s not a lot of ego.”
After eight years at Google, Girouard seems to be loving the startup vibe. He likes to quote Andy Berndt, head of Google Labs, who used to say (paraphrasing here), “Why is it that 10 people sitting in a room at a startup is bliss and 10 people sitting in a room at a meeting is painful?”
There is no indication at all that anything about Upstart is painful to its founder. “The funny thing here is we’re all in a big room with four big tables, but no matter what happens, we all end up clustered around one table,” Girouard said.
Also from GigaOM: