During the dot-com era, a new set of heros were discovered: startups. Rejecting the suit-and-tie of their former lives, brave men and women wore jeans and baseball caps to work and demanded comfortable chairs. They shed the shackles of their soul-stifling 9-5 jobs in Corporate America and worked long hours for smaller companies. Fortunes were made and lost in a matter of weeks. Porsche Boxters littered the California highway, the People’s Car of this revolution.

The story ended in tragedy, but we were all taken on a wild ride, and we enjoyed it. We are chastened now; investment money is harder to come by, and companies are often forced to figure out their product before they figure out their IPO. And yet the idol of the startup founder remains with us, working long hours in jeans and sneakers, drinking too much coffee, negotiating eight-figure venture funding deals. Dreaming of a 12-month concept-to-funding-to-sale cycle (or is it funding-to-concept-to-sale)?

This idol is back, in an upgraded 2.0 form. The names have changed (Digg instead of Lycos); they are often smaller and more frugal; and many of them will succeed.

But the idol needs to be dispelled.

Meebo is a startup in Silicon Valley that provides an in-browser, cross-protocol IM program. Think iChat or Trillian in a browser.

Business Week recently published an article on Meebo. It’s a typical dot-com profile. Meebo has 10 engineers, 7 business types, and is hiring. Their product is fairly complex and cool, with an Ajax front-end, a C++ backend, and a made-up name. Its developers work 14- to 16-hour days to keep the site running and to add new features like Meebo Rooms (kind of like chat rooms, but with 100% more Meebo). The environment is fun and informal, with ethnic take-out, games, toys, inside jokes, and Simpsons posters. Meebo also has serious VC funding from Sequoia Capital and Draper Fisher Jurvetson.

This makes for an exciting Business Week article: our eHeros are back! But two things give me pause.

First, 15 hour days are not a good thing. Long days at a startup sound sexy, like Aeron chairs and razor scooters. There are plenty of people out there that those kind of hours are ubiquitous for tech startups, like a company can’t succeed on 40-hour weeks. But death marches typically end with, um, death. Those hours can ruin a company and its employees. And more importantly, working long hours isn’t good for software. Software development is hard, and 40 good hours are better than 80 tired hours. I know developers who have burned out, gotten physically sick, and divorced because of this kind of thing. It isn’t healthy, sustainable, or profitable.

Second, Meebo won’t be a success unless it sells for $500 million, according to Business Week, and that just isn’t a good business plan. Paul Graham explains this well:

Venture investors like companies that could go public. That’s where the big returns are. They know the odds of any individual startup going public are small, but they want to invest in those that at least have a chance of going public.

Currently the way VCs seem to operate is to invest in a bunch of companies, most of which fail, and one of which is Google. Those few big wins compensate for losses on their other investments. What this means is that most VCs will only invest in you if you’re a potential Google. They don’t care about companies that are a safe bet to be acquired for $20 million. There needs to be a chance, however small, of the company becoming really big.

In other words, if VCs put $10m into Meebo in exchange for 40% of the company, then they’ve valued the company at $25m ($10m / .4 = $25m). So if they want a 20x return, then Meebo isn’t worth selling for less than $500m ($25m * 20 = $500m). In practice, they might be interested in cashing out at $100m if things look “bad” (for a 4x return), but they certainly wouldn’t want to sell for $20m, since that would actually represent a loss.

Maybe Business Week is exaggerating – I hope they are. I have no inside information on Meebo, and so maybe Business Week (or I) have the details wrong. But really, how many web startups sell for half a billion dollars? Maybe one a year? Off the top of my head, we’ve got MySpace, YouTube, and Skype (if Skype counts) over the last few years. Does Meebo provide anything as wide-reaching or revolutionary as these?

Venture capital may be great for biotech, aerospace, semiconductors, &c. But $10m isn’t needed for most web startups. There is another approach to startups, which understands the inefficiencies of large teams and the value of constraints. Throwing $200,000 at a programming problem often produces better results than throwing $2,000,000 at it. Not always, but often. If you have enough money to bootstrap a small team (3-5 developers) for a year, and if you (ahem) have a revenue model, you’ve got what it takes to build a business. You probably don’t need venture money.

And really, who wants to start a web business that isn’t a success at $30m?