(FORTUNE Small Business) – Stop for gas at any Exxon-Mobil, and you'll see Upromise logos featured prominently on every pump. Stroll for groceries at stores like Stop & Shop or Food Emporium, and you'll see that same logo--a small "u" wearing a graduation mortarboard--right there on the shelves. Mention Upromise (pronounced "you promise") in a group of parents, and chances are at least a few of them will have opened, or at least meant to open, an account.

Upromise seemingly came out of nowhere, and suddenly it's everywhere. As ideas go, the one behind it is so simple and smart, you wonder why no one ever thought of it before: You sign up for free, shop as you regularly would--buying things like gas and cereal--and a small portion comes back to you as a rebate. The money builds up in an account, and by the time your little scholar is ready for freshman year you'll have something to help foot the bill (see box "How Upromise Works").

For consumers the benefits are clear--this is money you'd be spending anyway, so you might as well get something back. And Upromise's corporate partners get something too: customer loyalty. About three million people have signed up so far, and founder Michael Bronner has lined up 145 of America's biggest corporations to participate--including McDonald's, Citigroup, Kellogg, and Kraft--as well as websites like Amazon.com and some 8,500 restaurants. AT&T gives you 4% on your long-distance bill. Coca-Cola kicks back 3 cents for every dollar spent on its products. General Motors will give you $150 if you buy or lease one of its cars.

Upromise has also generated attention because of the star power of its management team. Before founding the company, Bronner, 43, was a pioneer in so-called affinity marketing (he helped develop the American Express Membership Rewards program) and later launched Digitas, a marketing company that went public in 2000; he left with more than $200 million. Bronner serves as chairman of Upromise, and he hired George Bell, the former head of Internet portal Excite, as CEO. The board of advisors includes Kim Clark, dean of the Harvard Business School; George Fisher, chairman of Eastman Kodak; and even a Rockefeller.

But despite its management pedigree and the innovativeness of the idea (it's a graduate of our May 2001 Big Idea issue, and Bill Clinton sent the management a letter calling it "brilliant"), one major question remains: Can it possibly work as a business? After all, Upromise faces the same problems any upstart would right now: a spotty economy, a lagging stock market, and VC investors who will definitely want a return on their money in the not-too-distant future. In addition, as even its backers concede, Upromise represents an untested business model. "Can I point to a lot of successful for-profit companies that solely exist to meet a social need?" asks Mark Dzialga, a partner at General Atlantic Partners in Greenwich, Conn., which has funded Upromise from the start. "That would be hard to do." But that's the challenge for Upromise: Can a business that thinks like a nonprofit succeed in the for-profit world?

Considering how focused he is on higher education, it's ironic that Michael Bronner is a college dropout. His father died before Bronner's first birthday, so his mother, not yet 21, had to take on two jobs to keep the family afloat. Bronner had his first job by the time he was in third grade, at a local coin-and stamp-collecting store, and by fourth grade he had started a business washing the windows of apartment complexes.

However, it wasn't until college that he discovered his real business strength--marketing. Bronner was accepted into Boston University and paid for two years of tuition with a Veteran's Administration scholarship (his father had served in the Korean war). To cover the rest of his costs, he did what he knew how to do best: start another company. This time he put together coupon books, getting local pizza parlors and discos to pay to be included. The business took off. Bronner stuffed 100,000 mailboxes on college campuses all over Boston, and local advertisers were inundated.

He then took the idea one step further, persuading Gillette, John Hancock, and other Boston businesses to create similar books as an employee benefit. "For advertisers it was a no-brainer," he says. "It was a way to reach 125,000 customers." It was also the end of his academic career. Bronner dropped out of B.U. two classes shy of graduating. (He now sits on the board of advisors, though, and funds an entrepreneurial program.)

After leaving Digitas as a wealthy man in 2000, Bronner decided to apply his marketing experience to a more worthy cause. He launched Upromise that July and decided early on that he'd need an outside CEO. "I wanted to make a difference," he says, "not run an operation." To that end he brought in George Bell, a former documentary filmmaker and magazine publisher and the ex-CEO of Excite, an Internet portal with the distinction of having gone public faster than any company before it. (Excite went bankrupt half a year after Bell left, though a version of the site still exists.) Upromise had already raised $90 million from firms including Kleiner Perkins Caufield & Byers, Greylock, and General Atlantic Partners. The company began to market itself like mad--boosted by the fat budgets of its corporate partners--and the staff exploded to 220 employees (including one who shaved the Upromise logo into his hair).

Bronner originally considered launching a traditional charity, but then set Upromise up as a for-profit company. If it were a charity, he knew, every dollar he spent would have to be disclosed--and that covered not just paper clips and the staff foosball table but also salaries. In the midst of the Internet bubble it was tough to find talent, and Bronner knew that bootstrapping wouldn't bring in the best people or win the confidence of corporate partners. "If I did well, I could raise $100 million to give in scholarships, but that would help hardly anyone," he says. "Literally it would help 800 kids."

As a for-profit, his company can potentially help far more, but even Bronner acknowledges that no one's going to come close to paying Ivy League tuitions through Upromise. The company estimates that a typical account will accumulate $12,000 to $20,000 over 15 years. (As Bronner says, "It's a start.") That formula assumes, among other things, that you start with $60,000 in income and $12,000 in financial assets, and buy two-thirds of your gas at ExxonMobil, half of your toys from Toys "R" Us, and a GM car every seven years. It also assumes an 8% annual return over those 15 years, which is more than a little optimistic in the current stock market. Perhaps that's not worth quibbling over, though.

The bigger issue is how Upromise itself will make money. The company derives its revenue from a couple of sources. First, it gets a nominal fee for each customer transaction (for example, if you buy diapers, you'll get maybe 50 cents kicked back into your account, and Upromise will keep roughly a dime). The number of accounts already opened by customers is impressive--three million and counting--but there's at least some evidence that a percentage of those customers don't use it. Upromise won't say how many of the accounts are active or how much money is socked away in them.

Another revenue stream comes from the back end. Once your account hits a certain level, you're encouraged to put that into a state-sponsored college-savings program called a 529 plan (if not, it sits in the equivalent of a savings account). Upromise has partnerships with firms like Fidelity and Salomon Smith Barney, which offer 529 plans, and if you sign up with one of those firms, Upromise gets a fee. It may also collect a fee for any Upromise dollars directed into that fund, depending on the firm. Upromise won't say how many members have moved their money into these plans, but if they're like the rest of the population, it's not many. "Overall awareness of 529 plans is very low," says Whitney Dow, director of educational savings research at Financial Research Corp. in Boston. At least part of the problem is that stock market doldrums over the past few years have left people somewhat leery of investing for the future.

Still, these college-savings plans are a good deal--they're flexible, and they offer pretty enticing tax breaks. Dow expects assets in them to skyrocket from $15.8 billion in the third quarter of 2002 to more than $100 billion by 2005. And, understandably, Upromise is trying to capture some of that market. In November it partnered with Vanguard to launch its own 529 plan, the Upromise College Fund. To start it, the company raised another $12 million to set up expensive technology and hire new staff. Upromise collects a management fee for running that fund.

It's another good idea, and Bronner's heart may be in the right place (the Upromise plan lets you sign up online, and it offers a lower minimum than many plans). But the move means his company will probably end up competing with its own partners. Bronner acknowledges that they might not be pleased, but he says he wanted a plan that would be attractive to a wider demographic. "It's not publicly stated, but look at where they [the financial institutions] advertise and look at the minimums--it's not for people with low incomes," he says. "What about the people who are not financially savvy, who can't make those minimums?"

The still-private company doesn't release revenue numbers, but Bronner says it is on target to be profitable in mid-2003. That's good news, and something to warm the hearts of the VCs who've put more than $100 million into Upromise so far. Neither Bell nor Bronner would give details, but Bronner says those deals were fairly typical--the investors will want to cash out at some point, either through an IPO or sale of the company. Bronner says there's no time line, but Bell, the CEO, admits to feeling some pressure. "We think Upromise will be a financially successful company," says Dzialga, the venture capitalist at General Atlantic Partners. "That's why we made another investment in it in September."

"Everyone wants this company to win," Bronner says, and indeed he has qualities that make you want to root for him. He's taking a minuscule salary of only a few thousand dollars, and he's pledged all his stock in Upromise to an educational foundation (some of his management staff and investors have also pledged stock). As he says, "The problem with Upromise is that it almost sounds too good to be true." For his sake, we hope it's not too good to be successful.