Holden Karnofsky and Elie Hassenfeld were psyched. On December 20, their startup, GiveWell, scored a media hat trick, appearing in The New York Times and The Wall Street Journal, and on CNBC's Power Lunch. In a sign of the times, GiveWell was a nonprofit. Karnofsky and Hassenfeld, both 26, quit hedge-fund jobs last year and started a foundation with $325,000 from themselves and their friends. Their mission was to gather and disseminate exhaustive data on the effectiveness of charities.
GiveWell had found an important niche. American charitable donations reached nearly $300 billion in 2006, and charity is a classic long tail: 75% of that tally comes from individual donors. Unfortunately, most of it is dumb money. Large foundations have professionals who evaluate potential grantees, but their research is generally proprietary. Online resources such as GuideStar and Charity Navigator provide public ratings of nonprofits based on their IRS Form 990s but are skimpy on strategy or program details.
Hassenfeld and Karnofsky stormed the sedate corridors of giant foundations with brand names such as Gates and Hewlett and Ford, grilled big charities about their operations, and blogged about crashing philanthropy's "old boys' club." Tom Belford, coeditor of the Agitator, a blog that covers nonprofit marketing, wrote in a grudgingly admiring post: "GiveWell's audacity is breathtaking," calling Karnofsky "a 26-year-old punk, who doesn't know what he doesn't know."
Unfortunately, one thing Karnofsky and Hassenfeld didn't know was when to quit. Two weeks after their media debut, they were caught "astroturfing"—promoting GiveWell by using fake online handles and others' email addresses. GiveWell had championed authenticity and accountability, just what the sector and the giving public needed. Then, as board member Lucy Bernholz put it, "we fell on our face."
Bernholz, president of Blueprint Research & Design, a philanthropic-strategy consultancy, is one expert who has proposed a "social-capital market" in which donations can flow to the most effective organizations. But markets thrive on information—scarce in the nonprofit world, which has no P&L statements, no Standard & Poor's, no Yelp.com. "I have talked to so many people who say, 'We're going to create the Morningstar of the nonprofit sector,'" says Phil Buchanan of the Center for Effective Philanthropy, a consultancy for large foundations. "But the philanthropic road is littered with the carcasses of people who thought [applying business practices to nonprofits] was going to be easy."
It hasn't been easy for Karnofsky and Hassenfeld. GiveWell chose specific causes—to begin with, disease in Africa and education in New York—and offered a carrot: Nonprofits were invited to apply for a $25,000 grant. In exchange, the groups laid bare their operations. The two sought specifics relentlessly. On the cause of clean water in Africa, Hassenfeld says, he would ask, "Do you dig wells? Do you provide water purification tablets— and where, exactly, and how much of each?" All data were posted online with GiveWell's recommendations.
In a show of the transparency it hoped for from other organizations, GiveWell also blogged about its own internal processes. In a Christmas-morning post, after a marathon board session, Karnofsky noted, "People get tired and cranky," airing a knock-down debate over the organization's choice of grantees.
Just a week later, Karnofsky posted an abashed apology on the blog. He'd been caught posting questions and answers complimentary to GiveWell under misleading handles on several Web sites, including MetaFilter, and sending promotional emails to bloggers using another GiveWell employee's name and email address. Hassenfeld, too, had touted the group under a girlfriend's name. The board responded swiftly, demoting Karnofsky from executive director to program officer, and fining Hassenfeld $5,000.
"We need to hold ourselves to a very high standard of honesty, and what I did was wrong," Karnofsky told Fast Company in his first interview since the kerfuffle. At the time, he claimed a lack of sleep had led to his "lapse in judgment," an excuse that helped fuel a vicious pile-on in MetaFilter's comment threads, as did his offer to buy forgiveness by making a "donation" to MetaFilter. One commenter wrote, "It's nice to believe a couple of spoiled kids would chuck the Cristal and Lexus set and give their all for charity....But you just made it a lot harder to believe."
The worst thing about the GiveWell debacle is that it put at risk a service that's sorely needed in the nonprofit world. Credibility is hard to gain and easy to lose, and restoring it will be GiveWell's big challenge—tough for an organization aspiring to be an evaluator. As it starts a round of fund-raising this quarter, it's doing what it can to signal its commitment to accountability: Donor preferences will now help determine the causes it investigates. On a conference call with Fast Company, board member Tim Ogden said, "Like any nonprofit organization, our goal is not one of perpetuation of GiveWell. It's to alter a fundamental problem that we believe affects a lot of people." But Karnofsky, with a flash of the zeal that got him attention and then into trouble, interrupted. "At least I believe that GiveWell continues to have something unique to offer this problem," he said. "I believe that what we have to offer is going to be what's focused on rather than the mistakes we've made."
A version of this article appeared in the May 2008 issue of Fast Company magazine.