Tech trek Featured

9:34am EDT July 22, 2002

As Roger Byford, president and CEO of Vocollect Inc., demonstrates the Talkman, it's readily apparent just how good its voice recognition technology is.

He responds to prompts from an automated voice and the device captures his responses to "remember" how he says words and phrases. Vocollect's Talkman does it just as easily with Byford's British accent as it would with a Western Pennsylvania twang or an Arkansas drawl.

In a traditional warehouse setting, workers use pick lists, a printed list of items to be collected for delivery to a customer, to locate and select cases of merchandise for shipment. Typically, the employee identifies a location of an item in the warehouse by a number or combination of numbers and letters, goes to the location and selects the quantity specified. A label bearing information about the item -- price, size and warehouse location, for instance --is peeled from the pick list and attached to the case.

The Talkman eliminates the need for pick lists and stickers, relying instead on the operator's response to the orders the Talkman issues and the user's responses to those orders. Because the device is attached to the user's belt and employs a microphone and headphones, workers have both hands free virtually all of the time.

Vocollect's technology seems uncannily appropriate for this kind of application. It can be customized to the customer's particular use. Because the system requires no pick lists or labels, stray paper doesn't stick to

floors and equipment and warehouses stay cleaner. It eliminates annoyances like the difficulty in making labels stick to frozen food cases. And the check features of the system reduce selection errors and help warehouse operators spot merchandise misplaced on its racks.

Yet Vocollect endured no small amount of anguish trying to sell its technological wizardry, which initially targeted the manufacturing inspection market, to one set of users. It decided it had to identify another market if it planned to survive and grow as a company. But adversity and frustration pushed it to find new applications -- although not without growing pains -- for the clever technology it had come to master.

Vocollect ultimately found its market and adapted its know-how to it, but only after it became clear that it needed to change its focus and look for an easier sell.

Where's the market?

The marketplace didn't hear Vocollect's message nearly as clearly as the Wilkins Township company's Talkman comprehends users' utterances.

Much of the problem was unfortunate timing. The recession of the early 1990s hit its target customers hard. The final thawing of the Cold War was accompanied by the drying up of defense contracts.

Larry Sweeney, co-founder and vice president of marketing, explains how the company made the shift from the manufacturing market to the grocery warehouse market.

"Going from manufacturing inspection to warehousing, that's easy," Sweeney says. "The 1990-91 recession hit, and we laid off half the company. We needed to find a recession-proof industry."

Says co-founder Mike Gabrin, vice president and chief technology officer: "Things weren't going well in that time frame. We're selling to automotive, the first thing to go in a recession. The Cold War ends, we're selling defense electronics."

Vocollect's picture today is markedly different. It has found enough success to place among the 50 fastest-growing technology companies in Pittsburgh in the $5 million to $10 million annual sales category and is approaching the 100-employee mark.

Its growth forecast is rosy enough for the owners to commit to a new headquarters building designed to accommodate 300 employees.

A new focus

Vocollect hasn't abandoned the effort to apply its technology to the manufacturing environment, but it has scaled back after pounding doors and walking the shop floors of factories and plants for several years. Instead, it is concentrating on the warehousing market, customers which proved more desperate for a solution like what Vocollect had to offer, more willing to give it a try and more amenable to sharing their experiences with their peers.

But the shift didn't come until the founders realized they were failing to make an impact on the market they believed was perfect for the solution they could provide.

A solution looking for a problem

Byford, Gabrin and Sweeney worked for Westinghouse in the 1980s in the emerging field of voice recognition technology. When they left in 1987 to start Vocollect, they saw manufacturing inspection as a ripe market for voice recognition technology applications.

Quality control workers could inspect a part or an assembly and note potential problems that might indicate a problem upstream in the manufacturing or design process. A sag in the paint or a poorly fitting panel could be recorded and the data off-loaded to an information system that would collect it for analysis.

Problems in the process could be spotted early, allowing the plant to make adjustments or corrections before things got out of control. But while the manufacturing field could apply the technology -- companies including Ford, Saturn and Raytheon Corp. bought into it -- what Vocollect offered didn't provide (in manufacturers' minds) the clear advantages Byford and his partners thought it would.

"I think we were very much an organization that had a solution that we were trying to apply to problems," says Gabrin.

Byford recounts why the manufacturing arena, which originally appeared to be the prime application for their technology, proved so difficult to penetrate.

"You could sell to one Ford manufacturing facility, for example, and when you went to the one down the street or across town or in another state, it was another empire, almost autonomous, with another whole selling process to go through and a set of folks whose priorities might be completely different," Byford explains.

The managers at a Cleveland plant didn't much care what had been accomplished at their company's Atlanta site.

"It was difficult to replicate even within one company," Byford says.

And the competitive nature of the auto companies stifled any sharing of ideas within the industry.

Just as problematic for Vocollect was that companies in unrelated industries had little interest in what it had done elsewhere. Byford recalls an instance in which it tried to sell its technology to a company that produces nipples for baby bottles. To that prospect, Vocollect's track record meant nothing.

"The fact that we were doing things with Ford didn't mean a thing to them," says Byford.

The owners had promising technology that worked well, but not enough customers who found it sufficiently attractive to buy.

"We have customers who think the product we sell them is nice and does a good job, but when push came to shove, it wasn't mission critical," says Gabrin. "We wanted to find something that was mission critical."

Vocollect's partners learned a valuable lesson: They had a dazzling, inventive technology but weren't pursuing the strongest market for it.

"Just because you can do something doesn't mean you should do it," says Sweeney.

Seeing an opportunity

Gabrin visited a hardware distribution warehouse in St. Louis to see if it was a prospect for Vocollect's technology and came away a believer. The operations are complex, rely on large labor forces and have to meet tight deadlines for deliveries.

"It took you three seconds to watch them do their job, knowing the technology we had, to say there's an application," says Gabrin.

The warehouse environment proved a particularly good fit for Vocollect's technology for several reasons. First, it had the potential of cutting the cost of labor for picking, often more than half of a warehouse's total labor costs. It would eliminate case labels and allow workers to have two hands free at all times, thereby increasing productivity.

It could reduce the trash produced by the label backing paper and pick lists. And it could provide features such as spotting merchandise that was slotted in the wrong location or allowing pickers to go back to a previous item to double-check whether they had made the right selection.

Finding partners

About the time that Vocollect identified a new target market, some favorable events occurred. A major competitor in the warehouse market had been floundering and was sold off by its investors. And in the mid-'90s, high bandwidth radio frequency technology became available, making wireless transmission of large blocks of data more practical.

But while the market was a natural one, it also posed some challenges. In a manufacturing inspection setting, Vocollect's equipment could fail without a major disruption of operations. It wouldn't be so simple in a warehouse. The application would be central to a fundamental function in the warehouse environment.

"One difference on the inspection side was that it really wasn't mission critical and that falling back to paper and pencil was easy," Byford says. "The difference we faced in the warehouse was it was absolutely a mission-critical operation."

There were other technological barriers, as well. The hardware had to function in environments with wide variations in temperature and humidity. And it would have to be durable enough to withstand the bumps it would take in a warehouse situation.

Manufacturing proved a demanding environment for Vocollect's hardware, but the use by inspectors who worked at benches or walked around cars on an assembly line was nothing compared to the punishment meted out by warehouse workers.

"We thought we had a rough, tough environment in the manufacturing arena," says Byford. "We had no idea."

Vocollect set its sights on one of the largest grocery warehouse operations in the country to try to establish a foothold in the industry. It targeted Wal-Mart, the Bentonville, Ark., retail giant. Vocollect's top executives realized they had the software capabilities to meet the needs of the warehouse business, but recognized that with a small company, they needed to beef up their credibility.

To do that, they teamed with Telxon Corp., of Milford, Ohio, a well-known designer and producer of wireless and mobile information systems and a public company with 1999 revenue of $365 million, to design the hardware to be used with Vocollect's software. Together, they came up with the Talkman, a rugged but lightweight unit that mounts to the user's pocket or belt.

Wal-Mart proved a lucrative partner for Vocollect. In 1992, the retailer launched its supercenter concept, a combination discount store and full-blown supermarket, and was looking for ways to maximize productivity as officials implemented a plan to build five new distribution centers a year, each capable of supplying 100 supercenters.

With a company of that size, even small productivity improvements translate into large savings. As Byford points out, Wal-Mart's realization that no system would be absolutely perfect coming out of the box helped in the development of the product, which resulted in the Pick Manager software, which powers Vocollect's warehouse systems. By 1997, six Wal-Mart distribution centers were using the Vocollect technology.

The work for Wal-Mart has paid off. The company has systems installed in 16 of the retailer's distribution centers and expects to move into other Wal-Mart warehouses. It has sold into operations such as Roundy's Inc. of Pewaukee, Wis., a $2.9 billion wholesaler with eight distribution centers in five Midwestern states, as well as Kroger's and

The experience that Vocollect has gained with the warehouse industry applies to a very large market, one that goes beyond groceries to segments such as pharmaceuticals.

"A lot of these other areas look really interesting to us," says Sweeney. "There's a lot of volume there, and we can leverage our warehouse experience with these guys."

Obstacle turns into opportunity

What would have happened if Vocollect hadn't faced the problems it encountered in the early 1990s? It might have missed out on a golden opportunity, according to one of the partners.

"Our competitors would have found the warehouse market, probably before we did, and we would have been playing catch-up," Sweeney says.

While Vocollect has identified a niche in which it can leverage its success in grocery warehousing and gain advantages in similar types of operations, the founders believe that significant opportunities remain in manufacturing. Vocollect, says Byford, has a deep understanding of the technology, which creates a barrier of entry to potential competitors, and the technology has become more widely accepted.

This time, however, it will sell most likely through third-party systems integrators which will incorporate it into larger packaged solutions, and not directly to the end users.

"I think there's still a good application there, but I don't think it's the place to start a business," says Byford.

Perhaps not, but it may yet be a place to grow one. How to reach: Vocollect,; Telxon Corp.,; Wal-Mart,

Ray Marano ( is associate editor of SBN Pittsburgh.