Washed Up

Think brick-and-mortar companies have the answer for what ails the Internet economy? Guess again. Pandesic's painful lesson reveals serious cracks in brick-and-mortar attempts to reign supreme online.

Come February, the lights at Pandesic, the doomed ASP company created and then disassembled by Intel and SAP, will go dim once and for all, drawing to a close a costly Internet venture whose failure was most unexpected.

Just three-and-a-half years ago, the company launched amid great fanfare and expectation. It was going to demonstrate how brick-and-mortar companies would leverage their experience, clout and wealth to overtake juvenile start-ups that got a head start in building the new Internet economy. Pandesic aspired to be the first great application service provider (ASP) to offer a turnkey e-commerce solution. Instead, it collapsed after burning through millions of dollars.

Although its promise was far greater, Pandesic's fate was no different than those of the more than 130 dot-com companies that went belly-up last year, according to San Francisco-based market researcher Webmergers.com.

The lingering lesson of the Pandesic debacle is unsettling. It suggests that the brick-and-mortar giants being looked at to provide stability to an otherwise unstable market largely created by unproven start-ups are, in fact, as prone to failure as their flaky counterparts.

An Unstable Foundation
Pandesic announced on July 28, 2000, that it was closing up shop because it could see no way to make a go of its business. The company's downfall provides an invaluable lesson for others: Pandesic blundered because it tried building something grand on an unsteady foundation, which was supposed to support an unproven business model that was taken to market with a foolhardy approach.

In the old economy, companies were often hatched after great planning and deliberation. Saturn, the GM spin-off, is one example. The new car company took years of planning to launch, and combines the best of GM experience and resources with the radical new thinking and energy of a start-up.

In contrast, new-economy Internet companies are rarely born out of bureaucratic planning. One ingredient often found in successful Internet start-ups is two or more founders who are long-time friends or allies. For instance, Broadcast.com, the company that made NBA franchise Dallas Mavericks owner--and integrator--Mark Cuban a billionaire, was created by him and legal and accounting whiz Todd Wagner. They thought they'd make money broadcasting basketball games over the Web to forlorn sports fans who lived far from their favorite teams. Similarly, Yahoo, the company that bought Broadcast.com, was formed by two Stanford Ph.D. candidates, David Filo and Jerry Yang.

Pandesic, in contrast, wasn't formed by students or beer buddies eager to sacrifice everything they had to build a new-age company. A collaboration formed by two giants, Pandesic came together after SAP North American CEO Paul Wahl and Intel CEO Andy Grove decided the Internet could help streamline supply-chain management. SAP's R3 would serve as a foundation for a promising Web solution, the companies concluded, but it needed a front end. To make their dream a reality, SAP and Intel appointed Ed Harley, Bryan Plug and Harold Hughes to serve as Pandesic's founding managers. Although each man boasted an impressive resume--Harley, for example, helped transform former Ross Systems into a national powerhouse--together they were simply a collection of parts. As a team, they never worked seamlessly together, former insiders say. Plug, for one, left the company in early 1998 after a fallout with upper management at Intel.

In addition, the technology that served as the foundation for the collaboration was ill-suited to Internet time. SAP's R3, for example, requires numerous CD disks just to contain it all. Such a large code base was unwieldy compared with the slimmer, easier-to-manage solutions offered by successful Internet commerce-enabling companies, says former Pandesic director and current Vignette vice president John Vincze.

"The Pandesic solution wasn't written in Java or Visual Basic. It wasn't scripting-language friendly, so it put off most Web developers," he says. "It had operational excellence underneath and was great for cutting purchase orders on time, managing inventory and performing financial accounting flawlessly. But it wasn't suited to dot-com needs." In contrast to the SAP-based solution, Vignette's software can be downloaded via a dial-up connection, he adds.

Pandesic never fulfilled its dream of becoming a robust supply-chain solution. But it did create an interesting B2C catalog solution. Then it was one of only a handful of options. But by the end of 1997, there were several better-positioned alternatives in a market that some said had already peaked. What Pandesic's solution lacked were robust cross-selling capabilities, personalization and relationship-management tools, merchandising and upsetting options. With an unwieldy code base, adding such features would have taken years.

"Customers who used other packages could expect updates and enhancements in a matter of a few months," says Sheldon Laube, the former technology chief at USWeb, an early Pandesic business partner. "But the SAP code could only be updated every year to 18 months,an eternity in the early days of e-commerce."

One Seriously Flawed Model
In addition to its intricate management structure and its overly complex technology, Pandesic's foundation proved faulty for other reasons. For one, the company maintained separate facilities for former Intel and SAP employees who joined the alliance, creating operational teams that never coalesced. But the biggest obstacle to success was the unproven business model Pandesic tried to place on its shaky foundation. It literally broke the company, former insiders say.

Pandesic's software was always meant to be sold as a hosted solution, not a traditional, licensed software sale. At the time, that was pretty forward-thinking because it bucked the way companies traditionally sourced their IT technology needs. John Paget, CEO at GE Access, a $4 billion Boulder, Colo.-based products distributor that's trying to establish itself in the ASP field, says customers must modify their behavior before the ASP market can take off.

"We've had off-balance sheet leases for some time and have proven that we can manage that process effectively," he says. "But I do believe there is still a desire among many [customers] to be the owner of applications and infrastructure. That's the tough behavior-modification challenge the industry faces today."

No wonder, then, that market researcher Gartner Group expects 60 percent of ASPs to fail by the end of 2001. "Pandesic is only the tip of the iceberg," Gartner Group said in a report issued just days after Pandesic's decision to wind down operations. But what a tip. Unlike other ASP models, customers who chose the Pandesic solution were required to pay Pandesic a percentage of the revenue they generated by using the Pandesic hosted solution. The fixed percent that Pandesic wanted did the company in simply because customers hated it.

"The whole pricing model was crazy. It truly was a negative selection process," says Laube, now the CEO at CenterBeam, a Santa Clara, Calif.-based IT solution outsourcing company. The pricing model appealed to only the weakest and most risky customers. "Companies who thought or even knew they were going to be successful were never going to give up that kind of margin forever," Laube says. "So that left those who didn't know or couldn't know their fate."

Pandesic's decision to go with what amounted to a fixed-tax-per-transaction model only exacerbated customer resistance to rental-type agreements, he adds. "Such contracts have worked in other fields, such as the credit-card industry, but they've not achieved widespread approval in IT spending," he says.

As a result, Pandesic's customer base did not include the prestigious names the company had hoped. According to former employees, customers that bought in included Ouch.com, a company that sells health-care goods and KosherGrocer, a New Jersey business that sells Kosher foods online. Other customers included Vineyard.com, DVDexpress and HiFi.com. In all, the company attracted more than 100 customers, but only a handful of big names--and none committed to it significantly. For example, Adidas, the German sportswear company, signed up with Pandesic, but the deal covered only paraphernalia sold in conjunction with the company's short-lived World Cup Soccer site.

Square Pegs For Round Holes
Former executives of Pandesic say the company thought and acted big. Some wish it hadn't. "A cash-starved start-up recognizes very quickly when something isn't working and stops it immediately," Vignette's Vincze says. "But deep-pocketed companies think differently. Pandesic had a ton of money--and should have behaved like it had none sometimes."

Case in point: Pandesic staffers didn't sit on boxes or crates like workers at start-ups often do. Instead, workers sat on $700 Herman-Miller Aeron chairs. When they weren't sitting in their offices, employees were on expensive airplane trips. Granted, they flew coach, but they rarely opted for cheap, connecting flights, former insiders recall.

And Pandesic spared no expense when it came to branding and marketing itself. Prior to its launch, for example, it enlisted Lexicon, the company that named the Pentium microprocessor, to come up with a name. "Pan" stood for the company's global focus. "Desic," as in "geodesic," connoted the idea of an interconnected Web. All in all, the company spent hundreds of thousands of dollars on trademark research and other branding work. Not many successful start-ups do that.

Nor do many start-ups go forward with a meeting-oriented culture. But due to its two-office bureaucracy, Pandesic did so by necessity. Pandesic also blundered when it came to partnering. Diane Krakora, principal guru at Mountain View, Calif.-based Amazon Consulting and Pandesic's former channel programs manager, remembers aggressively recruiting partners only to have management undermine her work.

"I remember promising the moon to partners, only to see management quoted in news articles stating that the company was going to sell direct," she recalls. "It drove partners insane.".

Initially, Pandesic hoped to create a cookie-cutter solution that would make customer set-up a snap. On a typical $25,000 deal, Pandesic was willing to give third parties as much as $15,000 to connect customers to its solution. It figured partners would generate additional business providing custom front ends, templates and other Internet services. It planned to profit from its cut on each sale customers closed using the Pandesic solution. In reality, third parties found they couldn't connect the Pandesic solution profitably. Many said they wound up losing money on Pandesic deals because their service costs would far and away outstrip anything Pandesic shared.

Ultimately, Pandesic decided to build its own professional support department. But it did so in an up market when consultants, engineers and technicians were sought-after commodities and very expensive. So, Pandesic tried to build a professional services group at top dollar, but at the expense of potential allies. Pandesic soon found itself competing with the very companies it once hoped to recruit. The solutions they offered, which were built with software from Ariba, Commerce One, Intershop, Interworld, Microsoft, Vignette and others, were more flexible and less expensive. Pandesic found itself competing with square pegs in a market of round holes.

The great irony of the Pandesic debacle is that few appear to be taking notice. Instead, all eyes are back on the dot-com market--this time to watch as companies unravel. The lesson of Pandesic, however, should not be overlooked: Not all brick-and-mortar efforts are built to last.

"At the end of the day, Pandesic was simply overhyped," Laube concludes. "You had two businesses who knew nothing about running a service business--and it showed."

Mistakes Made Along the Way
Author, consultant and investor Geoffrey Moore believes companies often make the same, fatal mistakes. His books--which include Crossing the Chasm, Inside the Tornado and Living on the Fault Line--point out what smart companies must take into account before they can achieve long-term success. Some of Moore's best advice is embodied in colorful metaphors. Companies that fail to recognize the lessons from these metaphors often fail--Pandesic included.

The bowling alley: According to Moore, companies need to hit a few head pins to create positive momentum. In Pandesic's case, it expected world-class customers to beat a path to its door before it verified its business model or its technology.

The Escalator: According to Moore, companies tend to think of certain aspects of their business as core, when in reality they become commonplace over time, hence the analogy of running up a down escalator. Pandesic thought a powerful catalog solution would be core in Internet commerce. It was--for about six months.

The Fault Line: When faults slip or move, they do so quickly. Rather than adjust its unpopular model, Pandesic suffered damage when the fault line of e-commerce shifted. As a result, it never got to see how well it would do in Moore's tornado--the moment when everyone wants what a company has to offer.