Software Shakeout

Application service providers promised to transform the way business is done. What happened?

When USinternetworking Inc. (USIX) was launched in April, 1998, investors swarmed like yellow jackets around a honeypot. The Annapolis (Md.) outfit was a new kind of company--dubbed an application service provider--and it promised to transform the way software had been used by corporations for more than 30 years. Instead of selling customers large and complicated packages, USi would provide them with instant access via the Web to the software packages from such established software makers as PeopleSoft Inc. (PSFT) and SAP (SAP). No fuss, no muss. To venture capitalists and Wall Street investors, it seemed like a blockbuster idea. They bet nearly $500 million on the company.

Unfortunately, USi was almost as good at spending money as it was at raising it. By last October, only $60 million was left and the company was burning through a hair-raising $80 million a quarter. USi's stock, which had once topped $73 per share, had skidded into the single digits. The fate of the company rested on winning a $50 million loan from General Electric Capital Services Inc. It got it, along with $270 million in additional private financing from the likes of Microsoft Corp. (MSFT)

USi lives, but its near-death experience is symptomatic of an entire industry that popped up in its wake. Spurred on by analysts' projections that the ASP market would be worth more than $6 billion by the end of 2001, more than 500 companies were funded with a mind-numbing $10 billion in venture capital. The analysts, as it turns out, were wildly wrong. Concerns about the security of company secrets and the reliability of the Internet scared off many potential customers. At the same time, the ASPs faced huge costs for building data centers and licensing software packages from publishers. The result: Their money started running out before revenues kicked in. High-profile failures include Pandesic LLC and Red Gorilla, which both have ceased operations in the past six months, leaving more than 100 customers out of luck. Meltdowns like these put potential customers on notice to take precautions when lining up an ASP (page 80).

If this keeps up, the potential exists for techdom's biggest belly flop since the pen-computing fiasco of the early 1990s. The ASP market is expected to drum up only $600 million in sales this year, according to IDC, less than 1% of all information-technology spending--not nearly enough to support the horde of competitors. Market researcher Gartner Group Inc. predicts that 60% will go under in the next year.

TRUE BELIEVERS. By all rights, the money tap should have shut down for ASPs. But it just keeps flowing. That's because some of the leaders in the first generation--including USi--seem to be getting their acts together and a new generation of ASPs is being born. Rather than buying packages from established software companies, these new contestants, including accounting company NetLedger Inc., are building software from the ground up to run on the Web. That makes the services more efficient for them to run and easier for customers to sign up for and use. In the fourth quarter of last year, well after stocks of the few publicly traded ASPs had tanked along with the rest of the tech industry, venture capitalists pumped $1 billion into these companies. Crosspoint Venture Partners in December announced plans to invest $350 million in new ASPs that target specific industries, from groceries to banks. It's already backing 14 such upstarts. "This is the real future of software," says Managing Partner Rich Shapero.

Indeed, plenty of people still believe that the wave of the future is offering software as a service that's delivered over the Web. Corporations have been complaining for years about the expense and trouble of computing systems. They have to pay millions of dollars for the gear, then pony up again each year to maintain and update it. With application service providers handling these tasks, they can concentrate on running their businesses--and simply pay the ASP a monthly fee. Carey Eisenhower, Internet marketing manager for the Hershey Direct Div. of Hershey Foods Corp. (HSY), says he's saving at least 20% per year on software-management costs because he's a customer of USi. There is an added plus: "We didn't have the expertise to build an e-commerce site. USinternetworking did," says Eisenhower.

In spite of the early glitches, the basic ASP concept is compelling for software companies, too. Today, they depend on selling enough software packages each quarter to meet Wall Street's expectations. It's not unlike the movie business, where studios' fortunes depend on at least one megahit a year, creating spikes and valleys in their revenue streams. By delivering software as a service, their revenues should be more predictable. And when new technology is ready, it can be instantly included in the software and piped to the customer. While the first round of market projections were way off, researchers still see a sizable opportunity: International Data Corp., for instance, estimates $7.3 billion in sales in 2004.

LOOMING BATTLE. Those who like software dished up this way aren't going on faith alone. One company, VeriSign Inc. (VRSN), shows that offering up software as a service can work profitably. VeriSign, which sells encryption services for e-commerce sites and corporate communications, posted $474.8 million in revenues last year, up 460% from a year earlier. Pro forma net income hit $45.5 million. VeriSign is a winner because it spotted a technology needed by every company doing business on the Net, then beat others to the punch by offering it as a service. "Some companies may fail. But there's still an enormous amount of value to this market," says analyst William Martorelli of Boston market researcher Hurwitz Group.

That sets the stage for a bruising battle this year. With piddling revenue streams to be had in the short run, the early ASPs will fight desperately to get the formula right while next-generation companies pile into the market. And the old behemoths? Don't count them out. They, too, want a piece of the action. "The software business is going to change fundamentally in the next three to five years," predicts Oracle Corp. Chief Executive Lawrence J. Ellison. "Oracle is going to be ahead of that charge."

Some of the pioneers now seem to be on the right track. USi has rounded up 170 customers, and its revenues grew 208% last year, to $109.5 million. Analysts project revenues this year around $165 million, with the break-even point coming in the third or fourth quarter. The best news is that USi has finished building three huge data centers for running the software, which analysts say totaled more than $300 million. Another leader, Corio Inc. (CRIO), reported revenues of $43.6 million last year, up 650%. Analysts project $75 million in revenues this year and profits early next year. Both companies are now asking customers to pay part of the cost of software licenses up front. They have skirted problems encountered by some of the ASPs that failed, such as Pandesic, which tried to make money off of fees based on its customers' e-commerce sales.

No matter how well they tune their engines, though, the first generation ASPs will have a tough time outperforming the newer entrants. The newbies establish one super-reliable Web site that all their customers hook into--plugging their information into simple templates. They don't have to buy a new set of server computers for each individual customer, as the pioneers do. If the Young Turks get it right, the economies of scale could produce gross profit margins topping 90%--dramatically better than the 70% average among traditional software makers and the 15% to 20% margins that early ASPs have achieved, say analysts. Boasts Marc Benioff, chairman of sales-force-automation ASP "We are going to kill traditional software."

Don't think that big software companies will oblige Benioff. Microsoft, for instance, is betting its future on an online-service strategy. In addition to allowing ASPs to rent out such desktop applications as Word, Microsoft is building a technology foundation upon which other companies can build their services.

With a market as embryonic as this one, it's too early to call winners. The Microsoft of the ASP world might not even be alive yet. But, already, some of the contestants are promising. Analysts especially like VeriSign's prospects, since it has a proven track record and a dominant 75% share of the market for Internet encryption. And they're keen on Exult Inc. (EXLT), which delivers human-resources software (page 78).

While the notion of software as a service could turn the traditional packaged-software world upside down, the approach has deep roots in computing. For decades, companies such as IBM (IBM) have run other people's software from their data centers for a monthly fee. When the Internet came along, Web-hosting companies managed sites for tens of thousands of companies. USi's innovation was to offer the full array of corporate applications--from accounting to materials planning--as services delivered via the Web. IDC dubbed this an application service provider.

For a while, the computing world was nuts about ASPs. At the height of the mania, in the fourth quarter of 1999, venture capitalists pumped $2.5 billion into these companies. That was half of all the money invested in new software companies that quarter, according to McKenzie Consulting.

So what went wrong? The biggest obstacle has been inertia. It's just plain hard to persuade people to try something new. In interviews with more than 25 corporate customers, BusinessWeek found balky executives. Corporate IT departments are reluctant to give up control over their computing systems. CEOs are worried about Net security and fret about handing important business data over to another company, though those fears have so far proven to be unfounded.

MISSING LINK. For some, the numbers simply don't add up. "I just don't see the benefit," says Matthew Abraham, vice-president of information systems at Norm Thompson, an apparel catalog company in Portland, Ore. Abraham took a hard look at the ASP services of both USi and Corio because he didn't want to invest in the equipment and personnel to run new human-resources software from PeopleSoft. His conclusion: He would have actually spent 20% more per year with an ASP than if he handled everything himself. The ASPs claim they would have been cheaper partly because they believe Abraham underestimates the cost of building and operating fail-safe computing systems like the ones they provide.

When ASPs do manage to land customers, sometimes they fail to deliver the goods. Ben Young, owner of Big Blue Hat, a Web-design firm in Greenville, S.C., thought he had found a great deal when he signed up with Red Gorilla to keep track of his billable hours for clients. The service was cheap. But just as he was about to bill a client for a particularly labor-intensive project, Red Gorilla abruptly shut down because it ran out of money. "I guess you get what you pay for," he says.

Even among customers who get satisfactory service, there's a tendency to move cautiously. At Hershey, for example, just one tiny portion of the company, an e-commerce site called Hershey Direct, has gone online with an ASP. The rest of the company's computing systems are run through a separate computing division that sticks to running software the old-fashioned way.

Now ASPs are reconciled to slower growth than they had first expected. That's why they're focusing on profits. "The market is much different than it was a year ago," says USi CEO Andrew A. Stern. To cut costs, USi laid off 11% of its staff in January. Now it's asking customers to pay at least 20% of the total cost of a contract up front.

While USi and Corio are struggling to get on a winning track, they've got to be wary of the next generation of ASPs who are coming up from behind. In little more than a year, has landed 1,700 paying customers and 25,000 companies are participating in free tryouts., which provides small-business accounting services, has 3,000 customers. It's just a very different proposition than the one USi and Corio face. Since these newbies build their technology themselves, they don't have to pay a software supplier for programs.

It's low-risk for customers, too. While second-generation ASPs don't have all the bells and whistles of a mature software program from PeopleSoft or SAP, they make it very easy for customers to sign up and use the services. For NetLedger's service, the price is just $9.95 a month to start. A corporate customer can add more users at any time, and they also can drop them. "This thing saves me hours or days of work," says Michael Block, owner of Block Tax in Fort Lauderdale, Fla.

GROWING PAYROLL. The biggest challenge for the new crop of ASPs will be attracting big customers. NetLedger doesn't even try. It's meant for companies with less than 100 employees. Salesforce is aiming higher, but so far its largest customer has just 300 users of the service. Are these upstarts going to snatch away giant business software projects? Not yet. "Maybe in a few years I would consider working with them," says Nick Amin, vice-president of information systems at Cigna Corp. (CI) in Philadelphia. "They don't have all the capabilities that I'd need at this point"--such as sophisticated risk-analysis tools.

That's already starting to change. The upstarts are quickly adding such features as sales forecasting. And their deals are getting bigger, too. Employease, an Atlanta-based human-resources service, has seen its average customer size increase from 96 employees to 441 employees during 2000. "I'm supporting 500 people with Employease," says Reed Vickerman, vice-president of information technology at Copper Mountain Networks Inc. (CMTN) in Palo Alto, Calif. "Adding more would just be a matter of clicking a button."

The software giants can't turn on a dime like the upstarts, but they're quickly becoming forces to reckon with. The furthest along: Intuit Corp. (INTU), the $1 billion maker of financial and tax software for small businesses and consumers. Already, Intuit is reaping more than 20% of its revenues from online services. Every piece of packaged Intuit software, from the QuickBooks accounting program to TurboTax, has an online counterpart.

While Intuit has the jump on its brethren, other major software makers vow to excel at delivering their software as services. Software heavyweights such as SAP, Oracle (ORCL), and PeopleSoft have a distinct advantage over the first-generation ASPs when it comes to profit margins. They don't have to buy somebody else's software--they make it themselves. In the past year, Oracle doubled, to 100, the number of customers using its finance, manufacturing, and customer-service applications online. With SAP, 16 ASPs are hosting its applications for 150 customers. In early February, SAP said it had begun selling services directly to customers.

Microsoft is going at this market a bit differently. While it offers its Office desktop applications to customers through a handful of hosting services, its main goal is to provide a software platform for corporations and ASPs to build upon--the so-called .Net technology. As part of Microsoft's $50 million investment in USi last fall, the upstart agreed to offer customers services based on Microsoft's technology. But Microsoft faces stiff competition from Sun Microsystems Inc. (SUNW) and Oracle, which also supply foundation technologies for Net services. Upstarts such as and NetLedger built their systems on Sun's heavy-duty Unix operating system and use Oracle databases. Would they ever retool for Microsoft? "No way," says NetLedger CEO Evan Goldberg. "We don't think Microsoft software is ready for an online service."

Making the transition from the traditional software business to services won't be a snap for the established giants. They've got to retool their technologies to run smoothly on the Web. And they've got to tinker with their business models without upsetting quarterly earnings--shifting from their dependency on huge, up-front license payments to monthly fees.

Rather than mess with something that's working well, some of the established software companies are opting out of the ASP business--at least, for the time being. Consider Check Point Software Technologies Ltd. (CHKP), which sells firewall software that protects corporations from intrusions by thieves or hackers. It's deeply committed to a network of 1,000 distributors that handled 90% of its $425 million in revenues last year. Offer services directly to customers? "We could never, never do that," says Check Point CEO Gil Shwed.

Shwed, however, may be in the minority. Most established companies are gung-ho for offering software services. "If software companies don't do this--maybe not today but somewhere down the line--they are going to die," says Tim Chou, president of Oracle Business OnLine. That may be too dire a prediction. However, given the volatility of the software business these days, it makes sense for them to at least hedge their bets.

With Jay Greene in Seattle

By Jim Kerstetter

    Before it's here, it's on the Bloomberg Terminal.