Clothing label Karman couldn't find a suitable software vendor. So it became one

Three years ago, Gary Mandelbaum, CEO of Karman Inc., had good reason to believe that his latest line of Western-style clothing was a winner. Early word from the 20 sales representatives was that the $50-million, Denver-based company's new clothes -- especially a men's shirt with a bright Southwestern design across the front -- were selling like crazy.

That was the good news. The bad news was that when the orders were tallied, some three weeks after they were submitted, Mandelbaum realized that the company's inventory of shirts had been sold several times over. Many customers were going to be disappointed, and it would take weeks to meet the demand. What's worse, this scenario was becoming all too familiar at Karman, which was releasing a new line of clothes every 13 weeks.

The problem was obvious. Orders were taking so long to get processed that the company didn't know what the reps were selling. If Mandelbaum could see shortages as they developed, he could step up manufacturing or ask reps to hold off until inventory caught up with demand. He would also be able to spot slow movers so that he could let the sales force know where to push harder. "We just didn't know how much we were selling," he groans, "and by the time we did, it was too late to react."

Why the three-week delay? Because reps were handwriting orders on paper forms and shipping them to the company, where the orders piled up until a small army of temporary employees could enter them into the computer system. Equally troubling: 20% of the handwritten order forms were filled out incorrectly, and 10% of the computer forms had at least one keyboarding error. Most of the mistakes weren't discovered until the goods reached the stores, irritating customers and requiring time-consuming order audits.

The solution was equally obvious: the company needed a sales automation system. But after six months of looking, Mandelbaum hadn't found anything that seemed well suited to the garment business. Until, that is, he attended a Western-wear conference in Hilton Head Island, S.C., in 1992. One of the presentations was a 20-minute video by GO Corp. on its pen-based software operating system. The video demonstrated the ease with which pen-based computers could handle various chores in the field. Insurance adjusters, for example, could call up an image of a car, touch the fender to get the price of a fender repair, and then print out the repair cost and what the insurance company would or would not do, right on the spot. In the same way, pen computers could help police officers write tickets, doctors prescribe treatments, or . . . clothing sales reps take orders?

At least that's what occurred to Mandelbaum as he watched the video. Unfortunately, as he soon discovered, no such applications were available. GO was pushing a pen-based operating system, the basic software platform for pen-based computing. Someone else would have to write the application software. Well, thought Mandelbaum, how about us? Karman had some good programmers who could handle the job. And he could always hire others. Who knows, he thought, if the program turned out all right, maybe Karman could sell it to other companies in the garment industry.

Mandelbaum didn't know it, but he had just entered the software business.

* * *

Garment-industry salespeople tend not to be cyberjockeys, especially if they're in the 30- to 60-plus age range that tends to predominate at companies like Karman. They're not skilled typists either. "We noticed that very few people could work with a keyboard and talk or think at the same time," says Morgan Porter, Karman's vice-president of information systems. Porter had listened to Mandelbaum excitedly explain his Hilton Head epiphany, and they had brainstormed about the pros and cons of building a pen-based order-processing system.

The two quickly convinced themselves that the benefits of developing a system from scratch would be worth the risks. The decision was easy: both had come from a bootstrapping, no-nonsense background; both had grown up in a family business. From about age 11, Mandelbaum had spent every summer working at Karman for his father, until he'd mastered every aspect of the Western-wear trade. Porter had labored in the construction business for his father, doing industrial painting and sandblasting. Starting up a software affiliate to service first Karman and then other companies didn't intimidate either man. After all, if there had been any competition, they would have been able to find a system for Karman. And thus was born RuggedWare Ltd., a venture jointly owned by Mandelbaum and Porter.

Mandelbaum and Porter realized that they needed to make their computerized system work as much like the old system as possible. Sales is a game played with paper and pencil. Introduce a machine that demands focused attention, and you disrupt the delicate social balance between customers and salespeople. Salespeople should be schmoozing with prospective buyers, reading their reactions, and tuning into their needs, not trying to remember which key combinations they're supposed to type to get the screen menu they need. (Does Ctrl-F3 or Alt-F1 tell me whether the shirt comes in blue?)

In addition the ideal system would provide the latest inventory controls. Even the best computerized ordering isn't worth much if a salesperson can't get instant access to current information on pricing, inventory, sizes, color selections, and delivery times. There's not much point in making ordering more efficient if merchandise can't be delivered. And the system would have to slash error rates, too, by recognizing incomplete or inappropriate data.

Mandelbaum just knew that a pen-based system was the key to breaking his paper jam. He was amazed that software vendors didn't see it. He and Porter decided to get the operating system from GO immediately and put together a programming group.

There was just one hitch. It turned out that GO was using the video to attract investors. That's all there was -- just a video, no product. "I had never heard of anything like that," says Mandelbaum. "In our business you deliver -- very quickly -- what you show." But there was nothing usable from GO, and nobody else offered anything comparable.

Mandelbaum and Porter had to cool their heels for nine months before they could get a prototype from GO and start developing their first version of Pen on the Road, as they dubbed the envisioned software. But within months it was clear that GO as a company had pretty much, well, gone. Windows was too entrenched among software developers and customers for GO's operating system to get a foothold.

Although Porter and Mandelbaum were set back several months, they found that Windows could work with pen-based systems, and by 1993, Pen on the Road Version 2 for Windows was completed. It ran on a laptop computer equipped with a special pen for writing on the screen -- there was no mouse or track ball. With a bar-code scanner attached to the pen, a salesperson could swipe the pen scanner across a clothing sample tag, handwrite order information on the screen, and then pass the pen to a buyer to sign the order.

The first RuggedWare customer, of course, was Karman. But instead of thrusting the software at the entire sales force, Mandelbaum and Porter selected three reps for a brief pilot program. To put the software in its best light, they made sure that one of the reps had a fair amount of computer savvy. And to build credibility among the most computer shy of the sales force, they chose computer novices to be the other two pilot reps. The three pilot reps were trained for one week and sent out with their laptops for a high-pressure field test. They returned with glowing reports for the other reps. "We let them sell one another, rather than relying on us," Porter says.

The rest of the sales force was quickly equipped with laptops and given four days of training. "It was done in kind of an onion layer," says Porter. "First, we told them only what they needed to know to get their basic function accomplished, to write an order. When they were ready for more, we taught them how to get in and out of the system. The goal was never to teach salespeople how to use Windows." Once trained, the sales force was turned loose on the customer base in a sink-or-swim rollout. Mandelbaum claims that wasn't a big risk compared to the alternative. "We couldn't have gone through another season the old way and survived," he says. Not only did the reps survive, they prospered. Any resistance quickly evaporated.

Having conquered Karman, RuggedWare needed to win outside customers. The clothing business was the target market because it was the industry Mandelbaum and Porter understood, and it was the market the software was designed for. But the two also recognized that what suited Karman perfectly might be a rough fit for another company. So they made the software modular. A "general apparel footprint" module provided 80% of the basic functions needed by any garment company's sales force; a customization module tailored the system to meet company-specific needs for reports, graphs, and memos. Instead of marketing a generic package or having to write a new system for each customer, Karman can "tune" Pen on the Road to each customer's needs. Even the on-line help text can be customized to reflect a company's policies and procedures.

From their own experience, Mandelbaum and Porter knew that companies in the market for big software systems often are afraid . . . of time overruns, runaway budgets, and painful transitions. To lower customers' resistance, the two decided to guarantee a negotiated price and time frame -- typically about $8,000 per user, including customization, hardware, and training -- and 60 days from first meeting to delivery of the complete system. What's more, they would help customers set up a pilot program and would provide all the training. It added up to a big promise, but, after all, they had done it for their own company.

Mandelbaum and Porter demonstrated Pen on the Road at trade shows, but their biggest marketing tool was Karman's sales force. The two decided not to try to sell executives on the benefits of the system. Instead, they let Karman's reps spread the word to other reps. They wanted enthusiasm for the system to bubble up from the field. So far, the strategy has paid off: RuggedWare has already landed its first two customers and has placed 51 units in the field. Mandelbaum and Porter are projecting a breakeven year, with revenues of $1 million, and expect that figure to jump to a profitable $2.5 million in the coming year.

Now Mandelbaum and Porter are gearing up to increase RuggedWare's offerings. At the top of the list is a "marketing encyclopedia system": software to help gather, prioritize, and make use of the flood of information that salespeople encounter, which includes company voice mail, memos, E-mail, trade journal articles, and even tidbits gleaned from the Internet. Explains Porter, "Sifting through this stuff takes a tremendous amount of time away from the sales process." The system will work in much the way the Internet routes information of interest to particular salespeople -- by setting up "subscription lists" based on the type of information each rep wants.

After that, the two want to expand RuggedWare by selecting a second industry ripe for sales-force automation. To help provide the industry expertise the company will need to build a second general footprint, says Mandelbaum, RuggedWare will probably look for a partner.

But the plans could change. After all, what if someone shows Mandelbaum a video about 3-D virtual-reality goggles?

* * *

Peggy Noonan ( is a freelance writer based in Denver. She writes about science and health issues.