Around the same time, across the city in a modest Noe Valley flat, Matthew Flannery, a 27-year-old product developer at TiVo, was receiving long e-mails from his new wife, Jessica. She was a consultant in Kenya and Uganda with Village Enterprise Fund, a San Carlos (Calif.) microfinance organization. The 26-year-old described how she drove from village to village, evaluating the living conditions of business owners who had received $100 loans from the fund. "A little bit of money does so much,'' she wrote. What if, Matthew thought, he started a company that would let Americans lend money to struggling entrepreneurs in Uganda, enabling those entrepeneurs to bypass banks and loan sharks?
In March, 2005, Flannery mocked up a Web site, Kiva.org, which, loosely translated from Swahili, means "agreement.'' A month earlier, Larsen and Witchel had launched Prosper.com, a site that helps individuals loan each other small amounts. Both businesses are pioneers in microlending, a field born in the 1970s that is gaining prominence. Muhammad Yunus, who founded microlender Grameen Bank in 1976, won the 2006 Nobel peace prize for developing the idea of making small loans to entrepreneurs who otherwise might not be able to raise money. Entrepreneurs and philanthropists including Bill Gates and eBay founder Pierre Omidyar increasingly consider microlending an important strategy for lifting people out of poverty.
Kiva and Prosper share more than the goal of increasing access to capital. Both strive to keep operations lean, and both wrestle with a complicated regulatory environment. But Prosper is a dot-com, and Kiva is a dot-org. Prosper is a classic Silicon Valley baby, nourished by $20 million in venture capital and heavy-hitting backers including Accel Partners, Benchmark Capital, Fidelity Ventures, and the Omidyar Network. As a nonprofit, Kiva relies on donations and grants, which so far total $250,000.
The different models affect not only the superficial -- guess which company has slicker offices -- but also the fundamental. One key question: How will the benefits and challenges of each model affect its results? The issue is increasingly relevant for entrepreneurs whose businesses have an explicit social mission, be it protecting the environment or helping the disadvantaged. "Choosing between for-profit and nonprofit is becoming more of a common dilemma for entrepreneurs,'' says Alan Abramson, director of the nonprofit sector and philanthropy program at the Aspen Institute. "This is becoming an uncomfortable choice to make because they're trying to do both.''
Soon after their dining table epiphany, Prosper's Larsen and Witchel each kicked in $10,000 and hired a lawyer to see if their idea would fly with the U.S. Securities & Exchange Commission. In early 2005, they started offering programmers contract work and the promise of equity, and soon they had six employees. "They weren't doing it for charity,'' says the sandy-haired, soft-spoken Larsen. "There was no question we were going to get funded because of our track record.'' In April, Prosper raised $7.5 million from Benchmark and Accel. Prosper moved into a 2,500-square-foot office with furniture left over from a dot-com that went bust. Larsen brought his computer and office chair from home.
Larsen and his colleagues developed a site where those in need of cash can set up an account and list the amount they want to borrow and the maximum interest they will pay. Prosper's analytic tools assign each borrower a rating based on his or her credit record. Other Prosper users then offer a portion of the capital and bid on the interest rate. Those with the lowest bids become lenders. Last February, Fidelity and Omidyar Network invested $12.5 million in Prosper. "We didn't need the money,'' says Larsen, "but we wanted the brainpower.''
At Kiva, Flannery also worked his connections, albeit far less moneyed ones. He asked Moses Onyango, a pastor Jessica had met in Uganda, to gather $3,100 in loan requests from seven businesses, including a vegetable stand and a fishmonger. "The idea was that we would divide the loans into shares and sell them to the friends and family from our wedding invitation list,'' says Flannery. The loans wouldn't collect interest, but the shares sold in three days. Flannery collected the money through PayPal and wired it to Onyango, who then distributed the money to the entrepreneurs. Each succeeded in repaying the loan within the year.
Encouraged, the Flannerys organized a second round; this time, they invited the public to make loans. "I had this idea about sponsoring a business,'' he remembers. "That word was running through my head because I grew up sponsoring children.'' (His family donated to World Vision in the 1980s.) Jessica worked with Onyango to collect loan requests from 50 Ugandan entrepreneurs. Matt built a tool to create MySpace.com-like profiles for each one and sent out a press release. In November, 2005, two high-traffic blogs, Daily Kos and Boing Boing, featured the site. Within three days, 50 loans had been made for a total of $25,000.
After talking with his mentor, Bob King at Peninsula Capital, and his brother-in-law, venture capitalist Peter Cochran, with Vulcan Capital, Flannery decided to take the nonprofit route. "What we were doing was highly experimental,'' he says. "I thought that as a nonprofit I could get early support I probably couldn't get from a VC.''
Unlike Prosper, Kiva works only with entrepreneurs, which it finds with the help of microfinance institutions in developing countries. Those partners bring their clients to Kiva instead of to a local bank. Kiva posts pictures of the borrowers on its site and shows how each would use a loan. When a loan is made, Kiva transfers the funds to its partners. Lenders, who are usually repaid within a year, can read blog updates from entrepreneurs they've sponsored.
Last July, Kiva's six employees moved into an airy, 1,300-square-foot office in a Mission District warehouse with exposed brick and an assortment of secondhand furniture. Kiva also inherited its best talent from other places. Chief Operating Officer Olana Hirsch Khan came to Kiva in May after six years in sales at Google, and Premal Shah, who made his fortune as one of eBay's first employees, became Kiva's president last year. Shah and Flannery now earn about $3,300 a month, and Jessica volunteers while she finishes at Stanford Graduate School of Business. Says Flannery: "In the future, we'd like to scale this to the point where people are paid the salaries that will give them incentive to do the work.''
Since July, Prosper's 25 employees have inhabited 111 Sutter, a 9,000-square-foot space with a majestic wood-paneled conference room and a showstopping view of San Francisco Bay. The rent is more than seven times Kiva's. Prosper's employees could do quite a bit better as well: As part of competitive pay packages, they receive options that vest in four years.
In all but four states, Prosper has secured a lending license by meeting the terms of the Fair Credit Reporting Act. So far, Larsen says, default rates roughly mirror those at major lending institutions. Prosper receives 1% of the loan amount; lenders pay an annual servicing fee of 0.5% of the amount they lend. That money leaves Larsen, who is CEO, and Witchel, who is chief technology officer, free to concentrate on building the business and to add new features to the site.
That's freedom Flannery would like to have. He spends three-quarters of his time raising funds. He can't replicate Prosper's model in the U.S. because the legal expenses would be too much. Instead, Kiva plans to expand by introducing low interest rates to attract more lenders.
It is not surprising that Prosper's reach is so much broader than Kiva's. In its first seven months, Prosper's 100,000 registered users have made $22 million in loans, averaging about $5,000. Kiva's 15,000 lenders have made $1.2 million in loans, with an average loan size of $500. That translates into success for both companies, whose founders agree that there is room in the wide-open field for both models. Says Kiva's Shah: "Right now there are so few people doing this and it's so experimental, we need to help each other. If either model has a significant flaw, it could destroy the credibility for everyone." By Sonal Rupani, Anne Tergesen, and Bremen Leak