In doing so, of course, [Google] infuriated Wall Street by turning its back on the traditional form of underwriting -- in which a big firm rounds up the usual suspects (institutional investors), conducts a splashy road show, sets a too-low price that will ensure a big first-day ''pop'' in the stock, prevents anyone but its Wall Street friends from participating, and then charges the company going public an exorbitant 7 percent fee. From start to finish, it's an insider's game, which is precisely what Google found offensive. Although Mr. Hambrecht's firm did not conduct the Google auction -- the honor went to Morgan Stanley -- he was an adviser to Google, and helped ensure that individual investors were able to bid for Google's shares right alongside the big boys.

The Google auction received mixed reviews, but Mr. Hambrecht remains a passionate believer in auctions as a fairer, cheaper and smarter way to take companies public. In the auction method, there is no insider's advantage; anyone who wants to buy shares can go online and bid for them. Mr. Hambrecht's firm, which has pioneered auctions in the United States, did eight of them in the five years before Google came along. In the ensuing year and a half, it has brought another half-dozen companies public, including Morningstar, Avalon Pharmaceuticals -- and Traffic.com.

In retrospect, though, Mr. Hambrecht views the Google deal mainly as a lesson in how not to run an I.P.O. auction. There was so much about it that still bugs him. Google's board turned the auction over to Morgan Stanley. The big firm focused only on the institutions, and really didn't care about individual investors. With so much skepticism surrounding the auction, the institutional bidders were surprisingly tepid, resulting in an opening price at $85. Of course, that vastly understated the real demand -- and the stock closed at $100 at the end of the first day. The big boys got their pop after all.

WILLIAM R. HAMBRECHT took a seat next to me at a conference table in the hip San Francisco offices of his eight-year-old investment bank, W. R. Hambrecht & Company. He put his laptop between us, and opened a file that contained the details of an initial public offering for Traffic.com, which it had taken public five weeks earlier. ''I've gotten to the point where I don't even try to explain the auction method without my laptop,'' he said. ''The most powerful evidence that it works is the results themselves.''

Ah, yes, the auction method. You remember the auction method for initial public offerings, don't you? It had its big moment in the sun in August 2004, when Google used it -- or, rather, a modified version of it -- in its eagerly anticipated I.P.O.

In doing so, of course, Google infuriated Wall Street by turning its back on the traditional form of underwriting -- in which a big firm rounds up the usual suspects (institutional investors), conducts a splashy road show, sets a too-low price that will ensure a big first-day ''pop'' in the stock, prevents anyone but its Wall Street friends from participating, and then charges the company going public an exorbitant 7 percent fee. From start to finish, it's an insider's game, which is precisely what Google found offensive. Although Mr. Hambrecht's firm did not conduct the Google auction -- the honor went to Morgan Stanley -- he was an adviser to Google, and helped ensure that individual investors were able to bid for Google's shares right alongside the big boys.

The Google auction received mixed reviews, but Mr. Hambrecht remains a passionate believer in auctions as a fairer, cheaper and smarter way to take companies public. In the auction method, there is no insider's advantage; anyone who wants to buy shares can go online and bid for them. Mr. Hambrecht's firm, which has pioneered auctions in the United States, did eight of them in the five years before Google came along. In the ensuing year and a half, it has brought another half-dozen companies public, including Morningstar, Avalon Pharmaceuticals -- and Traffic.com.

BACK to Mr. Hambrecht's laptop. ''The company wanted to sell 7.5 million shares,'' he said. ''We estimated a price of between $11 and $13 a share. And look at what happened.''

I looked at the list of investors, and the prices they had bid. There were bidders in the $20 and $30 range, but not very many. Above $12.51, there were bids for only 509,000 shares. At $12.50, though, the number jumped to 6.7 million, which was still a little short of the goal. But at $12, suddenly there were enough bidders to sell 14.2 million shares. Fidelity was in the mix, but so were firms you've never heard of, like Luther King Capital Management. Here and there, I saw the account numbers of some individual investors. And they were all treated equally: because the offer was oversubscribed, everybody who bid $12 and higher got 53 percent of the shares they had asked for, at the $12 price. Everybody who bid below that number got nothing.

The auction wound up raising close to $90 million for Traffic.com. W. R. Hambrecht took 4 percent for its troubles. And the first day pop? It was nonexistent. The stock went up 15 cents that day. Though it has since fallen back a bit -- it closed yesterday at $9.76 -- Traffic.com's shares will ultimately rise and fall on the long-term success of the business.

And when you think about it, isn't that the way it is supposed to work? The purpose of an initial public offering has never been to enrich big, powerful investors with a guaranteed one-day profit. Contrary to popular belief, those 200 and 300 percent first-day price jumps, which so enthralled investors during the dot-com bubble -- that wasn't good. That was terrible.

The purpose of an offering is to raise as much capital as possible for the company, capital it can use to hire employees and roll out products. When an issue is underpriced, it means the underwriter has hurt the company that has hired it. And when an issue is purposely underpriced, it means that the underwriter is taking care of Wall Street instead of its client. Mr. Hambrecht, 70, is out to change that.

He wasn't always a rabble-rouser. When he was 32, Bill Hambrecht was a co-founder of Hambrecht & Quist, a West Coast investment bank that helped finance the technology industry, taking public companies like Apple, Adobe and Genentech.

Even after Mr. Hambrecht left Hambrecht & Quist and started W. R. Hambrecht at the age of 62, the auction method wasn't so much his cause as his market niche. ''When we first started doing this, I honestly thought that the abuses in the I.P.O. market were becoming so obvious that other firms would follow us,'' he said.

Instead, the big Wall Street firms dismissed auctions as a crude, untested method, not nearly as good at setting a price as a living, breathing investment banker. And a pattern emerged. Mr. Hambrecht and his team would be invited to make a pitch by an independent-minded chief executive who found the egalitarian nature of the auction appealing. Invariably, though, the venture capitalists on the board -- classic Wall Street insiders -- would push back, and Morgan Stanley or Goldman Sachs would land the deal instead.

If there is one thing that kept Mr. Hambrecht from becoming too discouraged, it was the advice he was getting from a board member, Clayton M. Christensen, the Harvard Business School professor and author of ''The Innovator's Dilemma.'' Mr. Christensen's expertise is disruptive technologies, that is, technologies or business models that have the potential to overturn the established order. As soon as Mr. Hambrecht began describing his method of taking companies public, Mr. Christensen said to himself: ''This is one of those. When conventional investment banks disparaged Bill, I said, 'Of course. It is very predictable. This is the classic pattern.' '' Wall Street was pooh-poohing auctions because they were such a threat.

Mr. Christensen advised Mr. Hambrecht to stick to very small companies that Wall Street didn't care about. ''Sooner or later,'' Mr. Hambrecht recalls Mr. Christensen telling him, ''some high-profile player will choose to use it, and that will be a turning point.'' Everyone at W. R .Hambrecht assumed Google would be it.

In retrospect, though, Mr. Hambrecht views the Google deal mainly as a lesson in how not to run an I.P.O. auction. There was so much about it that still bugs him. Google's board turned the auction over to Morgan Stanley. The big firm focused only on the institutions, and really didn't care about individual investors. With so much skepticism surrounding the auction, the institutional bidders were surprisingly tepid, resulting in an opening price at $85. Of course, that vastly understated the real demand -- and the stock closed at $100 at the end of the first day. The big boys got their pop after all.

For Mr. Hambrecht, the real turning point was his next deal. Thanks in part to the publicity generated by Google, W. R. Hambrecht was invited to give a pitch to executives at Morningstar, the mutual fund research company. Morningstar had already retained Morgan Stanley, but the company's executives found the idea of an auction so compelling that they hired Mr. Hambrecht. He insisted that this time his firm had to run the auction itself. Morgan Stanley refused to go along, and walked away.

And what happened? Morgan Stanley had originally told the company that the price range was probably $11 to $13. But the auction, which took place last May, showed that the real price was a good deal higher: $18.50. It sold 8.7 million shares. Although 82 percent of the investors were institutional investors, there was no favoritism. Fidelity -- the 800-pound gorilla of institutional investors -- bid for 2.2 million shares at $17.50 a share. Because its bid was too low, it got nothing. Since the offering, the stock has moved steadily upward; it closed yesterday at $40.93.

Still, Mr. Hambrecht's struggles haven't ended. The choke-hold Wall Street investment banks have on the process has eased only slightly. ''The Chipotle deal would have been perfect for us,'' Mr. Hambrecht said, referring to an offering a few months ago in which McDonald's spun off the Chipotle Mexican restaurant chain. It was a Morgan Stanley deal. ''The I.P.O. was $22 and the stock was at $44 after the first day,'' he said. ''They left almost $175 million on the table. How could they do that?''

As it happens, some people are finally starting to ask the same question -- at least belatedly. The unsecured creditors committee of eToys, the failed Internet toy retailer, has sued the company's original underwriter, Goldman Sachs, asserting that it's I.P.O. was so woefully underpriced that it hurt the company's chances of survival. (EToys' shares jumped from $20 to $76 on the first day of trading.)

Goldman is fighting the suit, of course, but it has also responded in another way. It now puts boilerplate language in its prospectuses that says, point blank, that it is ''not an agent or fiduciary of the Company or of any Selling Shareholder.'' In other words, if it wants to take care of its Wall Street friends at the expense of the company, it can.

That's some endorsement for the current system, isn't it? Give me an auction any day.