Bottom Line

How would you feel if one day people just stopped buying your most popular product? Software vendor Michael Edell got a taste of that feeling in October 1998.

That's when Edell began to notice a disturbing sales pattern. His company, jeTech Data Systems, based in Camarillo, Calif., created and sold labor-management software, which automates the administrative tasks -- payroll, time tracking, and project planning -- that come with managing a large workforce. At the time, Edell's company was losing deals, but not to rival vendors. Instead, sales prospects were forgoing software products altogether, preferring punch clocks and paper -- in Edell's view, inefficiency -- to the expense and hassle that came with buying and installing jeTech's software.

Never in 15 years had Edell, then 35, experienced such flat-out rejection. He knew something was up. "When you look at our markets, a solution like ours makes so much sense," he says. "It had to come down to the fact that we were somehow doing the wrong thing."

Meanwhile, in another realm of the software universe, the application service provider (ASP) model for selling and distributing software was emerging. ASP software vendors act like high-tech landlords, providing their customers with software over the Internet for a fixed monthly fee. Customers usually access the software, which runs on the vendor's servers or on those of a third-party provider, through their Web browsers. Besides giving customers more predictable costs, the ASP model spares them the headaches of buying servers, hiring IT staff, and enduring lengthy systems-integration ordeals.

Edell pondered his lost deals and concluded that unless jeTech began offering Web-based software, the company might well be doomed. This past January, jeTech became, the first software vendor to offer an ASP model in the labor-management niche, according to industry sources. By then, of course, few software vendors didn't have some sort of ASP plan. But back in 1998, ASPs were practically unheard of. These days, companies with an ASP approach are backed by venture capital and are officially hot. Research firms are also crowing about the ASP model. IDC, for one, believes the market for ASP software will balloon to $7.8 billion by 2004, compared with the $300 million it took in last year.

But for all the noise surrounding the ASP trend, what's rarely discussed is the dilemma small vendors like Edell face because of it. What happens when the product you've sold for years for millions of dollars suddenly has no prospective buyers, only renters? Obviously, your cash flow plummets -- at least at first. Using an ASP system, some vendors reap the same revenues they would have received as traditional software vendors, but the money is paid out over time. Most actually make more in the long run by structuring their leases not for a given period but for as long as the customer uses the product. Either way, there's a big cash crunch up front as the vendor adjusts from a diet of six-digit pie to monthly morsels.

Software vendors also have to determine whether they should switch whole hog to an ASP model or continue to offer customers the choice of buying their product outright. Offering both options gets tricky. How, for instance, do you structure incentives for your salespeople so that they will push the "cheaper" ASP model? And how do you explain to customers who've recently installed your software that the very package they've just paid six digits for and that required months to get configured is now available at rental rates on the Web, with minimal customization required?

Those were a few of the challenges Edell confronted as he transformed jeTech into The issue was not whether jeTech would change, according to Edell. "It was how do you finance, deploy, and market that change?" he says.

Michael Edell decided that unless jeTech began offering Web-based software, the company might be doomed.

Edell lists financing first for a reason: adapting software to make it deployable on the Web is pricey because of the personnel (costly programmers and Web designers) and gear (the latest servers and network equipment) needed to do the job. Most industry observers agree that it's nearly impossible for a small software vendor to switch to an ASP model without acquiring millions in outside funding, both to pay for the overhead and to provide operating capital until rental revenues ramp up enough to cover monthly costs.

At best, switching to the rent-based model would mean two years of flattened sales. At worst, if leasing never caught on with the large businesses that typically use labor-management software, the move would drown jeTech in red ink. Both scenarios were tough for Edell to stomach.

But Edell could at least feel confident that he had a key advantage when it came to seeking external financing: a track record as an experienced pre-Internet entrepreneur. Edell had built jeTech from barefoot beginnings in 1983 into a $14-million, 85-employee player. With customers like Corning and Hertz, tiny jeTech competed with national powers like $254-million Kronos and Simplex, which has revenues of more than $800 million. Edell consistently kept jeTech's margins near 20%. During the company's migration from DOS to Unix, in 1983, and throughout the early-1990s recession, jeTech stayed debt free, growing without outside funding.

Edell doesn't hesitate to invoke his status as an industry veteran, a profits-first type who needs a swig of Maalox when he hears about Web-based companies with free products and deep deficits. "I'm from the old school," he says. Complementing Edell's sturdy résumé is his decidedly sturdy appearance. With his full head of jet black hair, the energetic, six-foot-two Edell resembles not so much a grown-up techie as he does a seasoned corporate battler. "Customers don't like to buy from rookies," says Brad Jones, partner at Brentwood Venture Capital, which in November 1998 poured $7.5 million into jeTech.

Edell met Jones through jeTech's corporate attorney, Leib Orlanski, who'd already worked with the company for five years and had sat on jeTech's board. He had seen the company grow nearly 600% since 1994. Orlanski's firm had connections in the venture-capital community, and it was quick to make introductions for Edell (with an eye, no doubt, on doing the company's IPO if everything worked out).

Even with the Brentwood money -- which jeTech burned through in 13 months -- Edell found himself in need of a larger credit line. He inquired at his bank, but it was from the old school, too, and wouldn't expand jeTech's then $2-million line, which was based strictly on accounts receivable. Eventually, Edell landed a $6-million line at Silicon Valley Bank in December 1999, just as the $7.5 million from Brentwood ran dry. The line easily tided the company over until its next cash installment: $16 million from Lehman Brothers and Brentwood Venture Capital, which came in March.

For such a heretofore penny-pinching soul as Edell, raising and spending money at that rate might have seemed foolhardy. But the dollars involved in turning jeTech into were comparable to what other companies have burned through in doing their entire ASP makeovers. Art Williams, a director and analyst with Giga Information Group in Cambridge, Mass., says that Edell's figures are not only typical -- they're necessary. "You're giving up your front-end revenues for an annuity stream," he says. Williams estimates that it can take a company up to two years to recoup the front-end revenues it loses during the transition.

But that hardly spells doom for small software vendors that don't raise mountains of money, says Katherine Jones of the Aberdeen Group, in Palo Alto, Calif. Jones says some vendors take an outsourcing approach: rather than executing an ASP model by themselves, the vendors simply form a partnership with companies that already have an ASP infrastructure, such as Breakaway Solutions or NaviSite, to name two. Essentially, Breakaway and NaviSite license application software from vendors, host it on their own servers, and act as de facto value-added resellers (except in this case they're actually renting, not selling).

Edell, however, was intent on turning into a stand-alone ASP. He saw it as the company's best chance to close the gap on competitor Kronos (which had yet to offer an ASP option). After all, the capital he raised could do more than just smooth over cash flow. It could also fund the staffing requirements for rapid growth, allowing the company to add salespeople, programmers, developers, and security gurus. So far, in fact, 85% of's capital has gone toward hiring personnel.

But the hiring hasn't always gone as planned. The company has actually had two separate recruiting binges in the past two years. The first ramp-up -- bringing the total number of employees to 180 by the end of 1999 -- was followed by a "scale back" to 130 when Edell opted to outsource some support and implementation services he had previously planned to keep in-house.

The retrenchment was also prompted by what might be termed "technocultural" issues, workplace tensions created by technology changes the company had to make as it switched course. Since 1997, jeTech's programmers had written software mostly in Java code, with a Sun Microsystems-based architecture. Most members of one of the product-development teams saw no reason to change when the company converted its software from installed to Web-based.

But a faction of that team believed -- rightly, as it turned out -- that jeTech's software would run faster on Microsoft architecture. They also thought key parts of the system should be written not in Java but in the programming languages C++ and Visual Basic. When those workers persuaded Edell to scrap more than a year's worth of Java coding -- about $1 million in labor, Edell guesses -- some of the Java programmers revolted and left the company.

Other programmers had grown accustomed to an environment in which they sometimes spent more than a year developing products for which time-to-market rates weren't crucial. Now the treadmill was accelerating. The structure of the product-development segment was changed, and the company's development teams were whittled from 20 members to 5, yet the teams -- faced with the same workload -- were expected to produce faster. "We took 18 months of work and crunched it into 3 for a development cycle," says Dave Mikelonis, vice-president of product development. Mikelonis says the company hardly misses the departed workers. "In my opinion, 20% of us were doing 80% of the work anyway," he adds.

Edell saw turning his company into an ASP as its best chance to close the gap on the competition.

Edell estimates that he lost 50 employees, between the disgruntled Java programmers and workers who resigned or were fired for their inability to adjust to the faster pace. And like Mikelonis, Edell thinks the business is better for the exodus. "No one left that we didn't want to leave. We lost 50 people, and production doubled," he says. Whether the jettisoned workers share Edell's view of their inefficiency is another matter. Edell refused to provide contact information for his ex-employees, saying: "I don't think trying to find out why they were left behind is appropriate. I don't want to rub that in anyone's face."

Although the ASP ramp-up began nearly two years ago, the company didn't change its name to until January, five months after it began pushing the ASP model. The belated switch symbolized Edell's biggest regret to date and another cause of employee defections: poor communication with the rank and file. When the company was still known as jeTech, confusion reigned among the staff. The ASP offering seemed like just another new product from jeTech -- not a complete change in corporate direction.

Edell faults himself for not spelling things out more clearly. In hindsight he wishes he'd taken employees on a three-day "vision" retreat. Instead, he gave a PowerPoint presentation to his staff to outline the ASP plan. He thought his vision was clear, as did his fellow executives. But the rest of the staff members were baffled by this new ASP product called that, according to Edell, was going to change the world. "There was a lot of confusion," says longtime salesman Dan Auge.

When salespeople demonstrated an early version of jeTech's ASP product in the field, they didn't see what all the hype was about -- and neither did some of their customers. That wasn't surprising since the early version looked just like the installed jeTech product. It hardly seemed worthy of the fanfare Edell had sounded in its honor. "They saw some product getting released that looked just like our old product, and they'd ask, 'What's the difference?" says Edell. "They didn't get it." Even after numerous informational meetings, sales staff would return from the field wondering what, besides the cost, were the benefits of a Web-based product. "It was hard for me to fathom why a company wouldn't buy the product the traditional way," adds Auge.

When the sales force finally did understand -- and start to sell -- the company's ASP solution, another problem emerged: how could the company motivate its salespeople to sell a three-year contract instead of a flat-fee installation? The sales staff was selling both versions at the same time but had little rationale to push the rental model, since doing so would mean smaller up-front commissions. Edell and the sales staff ultimately agreed on a compensation system in which the company would pay full commissions during the first six months of an ASP rental contract. (At press time Edell said he was considering paying them over 90 days.) That way, salespeople wouldn't have to wait three years to collect the commissions they previously would have received instantly on a sale.

Though Edell takes the blame for the communications gaffe, he attributes his company's bumps partly to its early entry into the ASP arena. In late 1998 there essentially was no ASP market, which placed a big burden on the company to explain the concept to both employees and customers. But the employees, it seems, finally do "get it," and has sold 20 Web-based deployments so far.

In addition, the company has yet to face any wrath from its legacy customers, who paid for on-site software installations before the company began offering its Web-based alternative. Edell doesn't anticipate problems, mainly because he's been upfront with customers about the switch, and because most of his customers seem to understand how rapidly technology changes.

But not everyone thinks it'll be easy for companies making the ASP switch to keep existing customers from feeling cheated. John Witchel, founder and CEO of San Francisco-based Red Gorilla, which offers Web-based time-tracking software, thinks established companies are at a decided disadvantage in rolling out an ASP model. "You have tons of customers who've spent millions on software. If you offer the exact same thing on the Internet for less, you'll have an absolutely livid customer base thinking it got ripped off," he says.

Other vendors, however, agree with Edell that customers are numb to the pain of frequent high-tech spending and are therefore unlikely to become angry when they have to spend more on new technologies. "When customers make an investment to build a data center, hire IT staff, and buy servers, they know that after a few years they'll need new stuff," says Allie Rogers, CEO of $10-million Triple Point Technology, based in Westport, Conn. "Over time, customers expect more users or volume on their systems anyway, which always means new purchases. There's no concept of making an investment and sitting on it." Though Triple Point's latest releases are deployable on the Web, Rogers says, customers are hardly champing at the bit for his company's Web-based products. "I still consider Web-based software two years away from wide acceptance," he says.

Jim Kizielewicz, vice-president of marketing for Kronos, shares Rogers's view that the ASP model will take some time to take off. Kronos launched its ASP offering in June. "We've had an ongoing dialogue with analysts, and the market appears to be extremely overhyped right now," Kizielewicz says. "Plus we weren't hearing about it from many of our customers."

The difference in opinion between Kizielewicz and Edell -- two guys in the same market -- illustrates the fallacy of generalizing about an industrywide trend like ASP deployment. Edell reacted to the model early for two reasons: he saw the technology's potential impact on his customers; and he realized that in his position -- as a small company competing with Kronos and Simplex -- he might miss the chance to capitalize on a historical change in the way customers purchase and use software. Right now it's difficult to judge whether is too early, right on time, or just plain wrong. What's clear, though, is that Edell is no longer quite the same old-school guy. "Well, now I'm kind of an old-school, new-school guy," he says. "New school when it comes to the model. Old school when it comes to making money."

Ilan Mochari is a reporter at Inc.

Please e-mail your comments to