(Business 2.0 Magazine) -- Talk about a killer app. Two years ago Jia Shen and Lance Tokuda wrote, just for fun, a goofy Web application for MySpace that could turn anyone's photos into live-action slide shows. It succeeded - horribly. Within days of its launch, hordes of users at the then-superhot social network discovered the app, added it to their profiles, and communicated it to their friends. It spread like a case of Ebola at the Super Bowl. Within a month Shen and Tokuda had 100,000 users, and traffic was doubling every 24 hours.

The servers - those digital canaries in the mine shaft - crashed, and crashed again. "It was crazy," Shen says. "We were down 17 of the first 30 days." Then it got worse. With traffic peaking at 1.5 million users, server costs topped $20,000 a month. And there was no way to monetize their creation.

The gatekeeper: Facebook platform manager Morin is busy keeping tabs on more than 2,500 applications.

Drawing power: Kantor (center) and brothers Tim and Ted Suzman turned to advertising to monetize Graffiti's 5.9 million users.

Mini moguls: Goldstein and partner David Gentzel run SocialMedia, one of the biggest startups dedicated to Facebook apps.

More from Business 2.0

Live chat: your new online salesperson The hijack-proof truck Server farm goes solar

Fastest Growing Tech Companies Current Issue Subscribe to Fortune

type="text/javascript"> cnnad_createAd("636629","http://ads.cnn.com/html.ng/site=cnn_money&cnn_money_position=220x200_ctr&cnn_money_rollup=technology&cnn_money_section=quigo¶ms.styles=fs","200","220");

Still, they soldiered on for more than a year, keeping afloat with tens of thousands of dollars in loans while hoping to figure out a way to turn their enormous fan base into a brilliant business. It never happened - at least not on MySpace.

This spring, however, Shen and Tokuda spent a few days porting their MySpace hit over to Facebook. The upstart social network began as a hangout for high school and college students and last September allowed anyone to join.

Eight months later, Facebook did something MySpace still hasn't done: It opened up its network to developers and made it easy for them to make money from their applications. Which is exactly what Shen and Tokuda did when they rewrote their app and let it loose on Facebook.

Two months later, the duo had generated more than $200,000 in ad revenue. By late July they had 14 other apps up and running, with more than 22 million users. "When we started, we had no idea what we were doing," Shen says. "Now we have a whole suite of applications, and that's where our power is."

It's an increasingly common tale as the Facebook economy picks up steam. In just 10 weeks, hundreds of developers launched more than 2,500 new applications, triggering 139 million downloads. While a possible Facebook IPO or acquisition could change things overnight, for the moment it's a free-for-all.

The apps have names like FoodFight, Zombies, (fluff)Friends, and Fortune Cookie, and they let users indulge in everything from scrawling graffiti and sending virtual cocktails to buying music, brokering loans, and joining charitable causes - usually without leaving their Facebook homepages. Some apps have attracted hundreds of thousands of users, and a select few have pulled in millions.

One venture capital firm, Sand Hill Road-based Bay Partners, has set aside more than $12 million to bootstrap 50 new Facebook applications. "The current apps only scratch the surface of what is possible," says Salil Deshpande, a partner at the firm. "We're looking for much more sophisticated applications that can make money."

Developers like Shen have already proven how to use Facebook - and other social networks - to pull in a mass audience. But figuring out how to profit from those viral applications is another matter. So far, most of the revenue from Facebook apps comes from fairly primitive forms of advertising, such as Google (Charts, Fortune 500) AdSense. Yet a few developers are building applications to sell real and virtual goods. And others think they'll be able to charge major brands for access to the highly targeted Facebook crowds they've started to assemble. "We intend to build a giant company on top of these social operating systems," says Slide CEO Max Levchin, who's already made one fortune as a co-founder of PayPal. His startup specializes in photo slide shows that pull in more than 129 million users a month. "It's an opportunity for all of us to build the next Electronic Arts, Intuit, or Adobe."

The Facebook economy was born one afternoon in May, when the insouciant boy hero of social networking, Facebook CEO Mark Zuckerberg, told a crowd of developers in San Francisco what they had been dying to hear: that hackers deserve a real piece of the action in a market with ad revenue alone approaching $1 billion.

"Right now, social networks are closed platforms," Zuckerberg told the assembled entrepreneurs. "Today we are going to end that." That day Facebook began allowing programmers like Shen to build as many apps for Facebook's 32 million users as they could dream up - and to pocket whatever money they made doing it, with Facebook providing access to both the audience and the programming tools needed to draw them in.

Programmers talk about Zuckerberg and Facebook in the same terms they once used to describe Bill Gates and Microsoft (Charts, Fortune 500), so great are the power that social networks wield and the perceived stranglehold Facebook has on its growing audience. MySpace, by far the largest of these networks, with more than 100 million users, was the first to see them as software platforms, allowing users to customize their profiles by adding simple apps. But when it came to sharing revenue, MySpace held its cards close to its chest: it would quietly permit developers to make money only when their users left the MySpace network.

Zuckerberg has turned the MySpace business model upside down: Not only is he giving developers their own real estate within Facebook - both inside users' profile pages and on piggybacked application pages - but he's allowing them to make money from their apps any way they can, from ad sales to direct purchases of services and merchandise. For example, download iLike, an app that lets you sample and purchase music, and the developer gets a 5 percent kickback if you end up buying songs from iTunes or Amazon.com (Charts, Fortune 500).

To incentivize developers, Facebook is also breaking ranks with rivals by sharing crucial data - such as a user's age, interests, and friends - that enables more sophisticated applications. Zuckerberg also set up a speedy approval process that allows most programmers to load their apps into the network in a matter of days.

Josh Kopelman, a Philadelphia-based venture capitalist and investor in such startups as LinkedIn and Yapta, sees more users coming Facebook's way (ComScore reports Facebook grew 270 percent last year, while MySpace grew 72 percent) - and even more developers. "If you were a venture-backed Web startup," Kopelman says, "and had to decide whether to focus on a site that welcomed you in and let you keep 100 percent of the revenue you generate, vs. a site with a vague policy that doesn't let you generate any revenue, it's not even a decision. It's an IQ test."

The real IQ test, of course, is figuring out how to create an app that takes off and makes money. So what defines a killer Facebook app? Senior platform manager Dave Morin says the stickiest applications are those that tap into the "social graph." That's Zuckerberg's oft-quoted term for the web of connections between users and their friends. "Most apps are only interesting if there is much more content below that widget," Morin says. "It needs to take you someplace different, do something more."

Morin's favorite example is (fluff)Friends, which lets the user place a cartoon image of a penguin, pig, squirrel, radish, or other cute object on his or her profile page. People can pet it, buy a habitat for it (with fake dollars), even buy a real T-shirt (with real dollars) bearing the virtual pet's image. That's all pretty standard stuff these days.

But this app's clever twist, Morin says, is the way it gets you to reach out to your friends. First you adopt a pet and invite your friends to pet or feed it. Then you pet their pets, or see all the pets that your friends have adopted within Facebook - all while racking up virtual currency to engage in more (fluff)Friend silliness. "I call it interaction capital," Morin says. "The more users interact with the application, the more virtual credit they get." And if they sell a lot of T-shirts or advertising, that's more cash in the (fluff)Friends creators' pockets. The app cleared 1 million downloads after just seven weeks and now adds more than 10,000 new users a day.

There's a science to achieving perfect viral alchemy, and it's getting more sophisticated by the day. One place every Facebook developer frequents is Appaholic.com. Created by San Francisco-based programmer Jesse Farmer, Appaholic breaks down Facebook applications by popularity, growth rate, and even "virality," as measured by growth in a single day. On a recent day, Farmer ticked off the leading app in each category: Top Friends, a Slide application that lets you rank your friends; Griddle, a word game; and What's My Chinese Name?

Appaholic has developers glued to the site's analytic tools, looking for secrets that reveal what makes one app soar and another tank. When a new feature suddenly boosts an application's number of users, "you quickly see other developers rolling out similar features," says Paul McKellar, the San Francisco-based programmer behind the hit app SocialMoth, with more than 400,000 users. "You have to, if you want to keep up."

In the short time since the new Facebook platform went live, Farmer has already spotted a few telltale patterns. One attribute that's death to an app, he says, is complexity. Facebook and all its homegrown applications are relatively simple; those who create something that requires too much thought or explanation quickly run into trouble.

Farmer learned the hard way: Bookshelf, an application he helped develop, lets you list, share, and search your books, movies, music, and games. It went nowhere. "We were decimated by applications that didn't do nearly as much, that were far simpler, like iRead," Farmer says. "Within a week they were 10 times our size. Any application that is more complicated than the most complicated feature in the core of Facebook will be penalized."

Applications that augment or mimic existing features on Facebook - such as the wall (a space for writing messages) or a poke (a way for friends to say a quick hello) - are also more likely to take off. And those that stumble on even the smallest bug are likely to become roadkill. Matches, a flirting application, fell into a hole when a time-out bug, a Facebook glitch, stopped the app in its tracks. In the week it took to fix it, Matches lost about 100,000 users and ceded the category to a rival called Crushes. The lesson, Farmer says, is "users don't care why it doesn't work or whose fault it is. They will leave and probably not come back."

Armed with those sorts of insights, some startups are positioning themselves as Facebook app factories. "Netscape browsed the Web, Yahoo organized it, Google searched it, and now Facebook has made it social," says Seth Goldstein, co-founder of SocialMedia, a small shop in Mill Valley, Calif., that's already turned out such Facebook hits as FoodFight (throw a virtual lobster at your buddy) and Happyhour (send that buddy a cocktail). How does he plan to cash in on all those widgets?

At the moment, advertising opportunities are unproven - which is why Goldstein is leaning toward sponsorship as a simpler path to profits. FoodFight, Goldstein says, is an ideal mechanism for food companies to market themselves. Instead of throwing a chicken drumstick at a friend, a user could throw, say, a drumstick sponsored by Tyson Foods. "I had an ad agency representing a buffalo wings chain approach us with an $80,000 ad buy," Goldstein says. "It's starting to happen."

Shen and Tokuda's outfit, meanwhile, has become a lot more than a slide show. The company, now called RockYou, has more than $10 million in venture funding, more than a dozen developers, and one of the largest portfolios of applications. Its 15 apps include Horoscopes, Emote (icons for your status box), and Glitter Text (sparkly fonts). This time around, the revenue model is getting as much attention as the code. In late July the startup launched its own advertising network: RockYou is offering its user base and Facebook pages as a way for advertisers and other developers to reach more users. "We don't know which approach is going to work best yet," Shen says, "so we're trying them all."

So is San Francisco-based Slide, which has 12 Facebook apps and a growing audience to offer advertisers. Slide is also launching an ad network that will let advertisers brand its apps. CEO Levchin thinks that because users volunteer their ages, interests, locations, and other specific personal information, Facebook has the potential to be the best ad platform on the Web. "Until recently, Facebook had all of this ad inventory to itself," Levchin says. "Now it's saying, 'Go nuts. Sell it any way you want.'"

Not everyone is drinking the Kool-Aid. Andrew Chen, an entrepreneur-in-residence at Mohr Davidow Ventures, thinks the revenue opportunity is still unproven. "The question is whether large-brand advertisers will feel like it's a good idea to buy space on still relatively small pieces of real estate," Chen says. "I would imagine they'd want to deal directly with Facebook." The company, after all, already generates an estimated $150 million in ad revenue on its own.

Developers face other risks: Should Facebook go public or get acquired - as has been widely rumored - new circumstances could force Zuckerberg to give up his share-the-love revenue model and keep more of it in-house. The company might also rip a page from the Gates playbook and launch its own versions of the most popular applications. Or Zuckerberg could kick everyone out and go home.

As a hedge, developers aren't limiting themselves to one platform. Bebo, LinkedIn, MySpace, and several other large social networks have signaled in recent months that they will likely follow Zuckerberg's footsteps. "They will all open up," says Charlene Li, a marketing analyst at Forrester Research. "It's inevitable." MySpace, for its part, is said to be working on substantial changes to its platform. While company officials declined to respond to specific questions about its plans, they did say their goal is to work more closely with outside developers.

Anticipating that day, Palo Alto-based Box.net, which sells online storage and sharing, recently created a Facebook app for its service and a subscription package for Facebook users. But that doesn't mean the startup won't be showing up on other networks when their doors open. While the networks all have different software protocols, the apps are small, and the time and effort required to retool one for, say, LinkedIn or MySpace doesn't scare developers. "Facebook has done the best job opening up," says Box.net CEO Aaron Levie. "But we are not about building a business on any particular platform."

For folks like McKellar, though, simply owning a few Facebook apps is just fine. He has yet to make any real money from SocialMoth, but he's willing to fork out $500 a month in server costs just to hold on to his audience in the hope that he'll figure out a revenue model soon enough. "I go where the users are, and where they make it easy for me," McKellar says. "Right now, that's Facebook."

1: Sell ads

The play

Just about any Facebook app can get into the ad game, but only those with the biggest audiences will earn serious money. Several easy-to-use ad networks are already delivering the ads for a cut of overall sales.(See "Tools," below.)

The front-runners

Graffiti (5.9 million users). This highly viral drawing tool spread quickly because of its simplicity and originality.

iLike (5.4 million users). Users can set up their music and video libraries in mere minutes.

The Simpsons Photos, Quotes, and Trivia (60,000 users). Pearls of wisdom from the first family of Springfield.

The payoff

Apps currently generate less than $1 for every 1,000 pageviews. But that amount will likely increase as demographic targeting becomes more refined and the ad models move from simply racking up pageviews to measuring users' engagement.

Tricks of the trade

1. Establish your base. Hold off on serving ads until you have at least 10,000 users. Bombarding users with too much advertising can scare them away and hurt your growth in the long run.

2. Test different ad networks. Putting up ads is a simple cut-and-paste operation, so you can afford to be choosy and pick the network that gives you the best deal.

3. Don't clutter up app pages. "This is definitely a challenge for developers," says Mark Kantor, one of three developers behind Graffiti. "The most important thing is to preserve user experience."

4. Renegotiate as you grow. Demand a bigger cut of the revenue share as your traffic jumps. Says Kantor, "It might be better to go with a small ad network if you think you'll stand out."

Tools

Dozens of ad networks are cropping up to serve the Facebook developers. Here are a few.

1. Lookery (lookery.com). This new Facebook-specific ad network aims to offer developers demographic profiles of their user bases. More targeted advertising could soon fetch a higher price.

2. Userplane (userplane.com). AOL-owned Userplane pays per minute of exposure rather than just per pageview, so it's good for applications like games that keep users highly engaged.

3. Google AdSense. Not new, but many developers consider it the best means of supplying relevant ads.

2: Attract sponsors

The play

Advertisers are already sponsoring apps. Besides being widely used, your application needs to offer companies a natural way to interact with their customers.

The front-runners

Likeness (2.9 million users). Offers quizzes that generate top-10 lists - an ideal branding vehicle - and matches them with those of friends with similar preferences.

FoodFight (2 million users). Virtual lunch money buys you food to throw at friends. Next up on its menu: chicken wings from a major food chain.

HotLists (1.6 million users). This app lets users define their personas by posting brands' logos, cleverly dubbed "stylepix," on their profiles.

The payoff

Building direct relationships with brands takes more time and effort, but it means higher-quality advertising and more control over how your users interact with it. Expect to earn multiple-dollar CPMs instead of the pocket change you'd get from the ad networks.

Tricks of the trade

1. Don't pitch big brands without big numbers. You'll need a large traffic base - at least a few million users - before top brands will pay attention.

2. Know who's looking at your pages and why. Analyze your user demographics so you can pitch your audience effectively to sponsors.(See "Tools," below.)

3. Let your users do the work. Incorporate brands that your users identify with, and they'll willingly spread the word.

4. Don't overdo it. Too much brand presence will scare away Facebook's sometimes advertising-averse audience.

Tools

Where to find help analyzing your traffic and users.

1. Google Analytics. Embedding Analytics into your apps is easy, and it churns out useful stats about where users are coming from.

2. Gigya (gigya.com). This startup tracks metrics like app stickiness and user adoption rates.

3. Appaholic (appaholic.com). This site tracks traffic growth by the hour, day, or week - critical when launching a new ad campaign.

3: Sell services

The play

As apps become more about utility and less about fun, opportunities will arise to sell digital services of lasting value to users. Eventually, they'll make purchases without leaving their profiles.

The front-runners

Files (43,000 users). Offered by Box.net, this online file-storage service turns a Facebook profile into a repository for members' digital media.

Picnik (206,000 users). A Facebook version of Photoshop.(Hello, Adobe?) Basic tools are free; advanced features are offered for an additional fee.

The payoff

If you're selling a real service, then you can have your cake and eat it too- try selling subscriptions and ads to double-dip on your traffic.

Tricks of the trade

1. Start with a free version. And make switching to a paid offering an easy process. Don't force users to leave Facebook to sign up.

2. Set logical limits. Decide carefully what you'll give for free and what you won't. And even the freebies must be valuable enough for customers to be willing to spend their time.

3. Research your price points. Box.net already had storage plans for businesses and professionals. But when it moved onto Facebook, the company rethought its pricing models and created a $25-per-year plan that's comparable to the cost of an external flash drive - the way most college students store important files.

4. Be tactful and timely. Box.net alerts its users when they're nearing their file or storage size limits, politely reminding them about its for-pay premium service.

Tools

Where to find a platform to process payments.

1. PayPal. A starter plan will cost you 2.9 percent plus 30 cents per transaction.

2. Google Checkout. The standard processing fee is 2 percent plus 20 cents per transaction.

3. Facebook. The company is rumored to be launching its own payment platform soon.

4: Sell products

The play

As Facebook increasingly becomes the center of people's digital lives, it's also becoming a venue for selling things - digital and otherwise - to its fast-growing audience.

The front-runners

Amazing Giftbox (127,000 users). Sends virtual Amazon merchandise.

Band Tracker (29,000 users). Searches upcoming concerts and links to ticket vendors.

Visual CD Rack (20,000 users). Lets users browse and buy music from a virtual CD rack.

The payoff

Most developers are going the affiliate route, offering product wish lists and then sending users to sites like Amazon.com or iTunes. Others, however, are directly selling such items as ringtones and T-shirts.

Tricks of the trade

1. Be a middleman. iLike makes its music-sampling apps simple and hands off sales to iTunes or Amazon via affiliate partnerships. Those directly selling hard goods need to prepare for the complexity of payment and delivery.

2. Keep it simple. Facebook has not yet become a place where people are likely to buy, say, a digital camera. But users are starting to purchase items that don't break the bank and extend Facebook's utility. XLR8 Mobile, for instance, is looking to sell ringtones and wallpaper on Facebook via custom storefront widgets. "We don't want to bring people to the store," says XLR8 Mobile CEO Perry Tell. "We prefer to bring the store to the people."

3. Give it away. Going viral is always the goal. One great way to get there is by offering free samples. Whether it's a digital download of a song or the image of an item, give your customers a taste of what they'll get before asking them to commit.

4. Don't rule out the odd. "Sometimes wacky, unusual, off-the-beaten-path stuff sells huge," Tell says. "Everyone is looking for the next Crazy Frog, so you must be willing to try lots of things."

Tools

1. Clearspring Technologies. This analytics service tracks exactly who's downloading an app and what they're buying through it. It also suggests when to double down on an item or sales approach that is working or, conversely, kill off those that aren't.

2. Garage Sale. Developers can use this Facebook shopping cart system run by Buy.com, which takes a 5 percent cut of sales.

3. Facebook Marketplace. The largest classified-ads community on the network, it's a good place to monitor buying trends.