Case Study 02

To bring his old-world company into the 21st century, this chocolate maker cooked up a homespun technology upgrade

Who: Bissinger French Confections
Where: St. Louis
What: High-end candy maker
Revenues: $6.5 million
Problems: Outdated technology; companywide inefficiencies
Solution: An IT "beauty contest" to explore the company's options

Ken Kellerhals remembers vividly his first Christmas as owner and CEO of Bissinger French Confections, a candy company in St. Louis that has been making its handcrafted chocolates in much the same way for more than 400 years.

Ruefully, Kellerhals refers to that holiday period as the "'96 fiasco." Through aggressive salesmanship, the opening of a new store in Minneapolis, and a refinement of his catalog mailing list, Kellerhals had managed to boost sales by 25%. But because he didn't have an accurate sense of his inventory, he was taking orders for more candy than Bissinger could actually produce. And because he lacked the ability to efficiently schedule his employees and the production of his various candies, he ended up running his plant for 18 hours a day and paying $189,000 in overtime during that period of three months. For a company with $2.8 million in sales, he says, "it was a disaster."

In the end Bissinger was able to fill the orders. But Kellerhals didn't have the tools to analyze his costs and figure out if in fact the company had actually made any money. The experience led to an epiphany. The CEO realized that the old-fashioned tracking systems and office procedures would have to go.

And so began the technological education of Ken Kellerhals, now 43, a man who admits that when he embarked on his upgrade adventure, all he knew was "how to turn the computer on and off." In his pursuit of technology, however, he was going to have to walk a fine line to preserve the heritage of the sweetshop's storied past.

The Bissinger family had begun creating its fine delicacies in 17th-century Paris and served as confectioners to King Louis XIV and Napoleon Bonaparte. Karl Frederic Bissinger was named "Confisseur Imperial" by Emperor Louis Napoleon. A descendant, also named Karl Bissinger, brought his recipes and chief candy maker to the United States in 1863, and his son (yet another Karl Bissinger) founded the company in St. Louis (where it now has four retail outlets) in 1927. The business stayed in the family until 1946 and then went through several owners before Kellerhals (who spent 14 years in banking as a commercial lender and 3 years in charge of operations at a toy company) and a group of investors bought Bissinger six years ago.

"We thought Bissinger was the best-kept secret" of the candy industry, says Kellerhals. Under the previous owner, he says, the company had been run as a small business, with no real growth plans and consequently no bank debt. But he saw the potential for growing the business, particularly through corporate gift programs and the company's catalog. To grow the company, however, he had to make it more professional, spend some money, and take a few chances.

One thing Kellerhals would not refashion, however, was the quality of his product. "We have a tremendous following of people who've been ordering from us for years," he says. "Our true competitive advantage is that the candies are handmade." To produce the candy by machine, Kellerhals explains, would mean changing the company's tried-and-true recipes, which in turn would alter the candies' taste. If you use a machine to place caramel in the center of a candy, for example, you have to first liquify the caramel, which permanently changes its consistency. Bissinger hand-dips solid caramel into chocolate and shaves off the extra chocolate when it hardens.

But just because the company was going to stick with copper pots and wooden spoons did not mean it had to continue using handwritten logs to monitor inventory and production. When Kellerhals first got to Bissinger, he recalls, the head candy maker at the 25-person plant kept a written diary of how much candy was made each day. The hardware that was in use, two PCs and 15 dumb terminals, could handle only data entry.

Ken Kellerhals, CEO of Bissinger French Confections: "We needed to look at recipes, at how much products were costing us."

The company did have three fairly decent software programs -- for order entry, shipping, and accounting -- that over the years had been customized to the business's needs. The problem was that the programs couldn't communicate with one another. That led to some inefficiencies. Price changes had to be entered into the three different systems, for example. And because the programs were not integrated, Kellerhals couldn't pool the data and perform the kinds of analyses he wanted.

"We needed a plant schedule," says Kellerhals. "We had to come up with some way to know what to make and when to make it. It's a tricky thing in this business because freshness is so important." For the company's inventory of some 900 products, the CEO wanted to generate a schedule for the entire year that would spell out what needed to be made when, broken down into months, weeks, and days. He wanted to compare production totals for each month with those for the same months of the previous year. With that ability in hand, Kellerhals hoped never to get caught flat-footed again, especially during the holiday season, when the company made 55% of its total annual sales.

Kellerhals knew he also had to start analyzing costs more effectively. "We needed to look at recipes, at how much products were costing us," he says. "When a nut crop fails, for example, it can have an enormous impact. Pecans can go from $3 a pound to $5 a pound." Without a thorough analysis of his costs and an understanding of how they affected his production and profitability, Kellerhals couldn't know the full effect of such increases.

"He had a clear vision of what he wanted to get out of his [computer] system," recalls Bryan Sapot, the consultant who eventually overhauled Bissinger's information technology. "He just wasn't quite sure how to get there."

Kellerhals's first step was to seek advice from associates and peers. He turned to some of his board members, some of whom had companies that had gone through a similar technological overhaul. He talked to hardware vendors like IBM and to contacts he had made in previous jobs. Mostly, he asked them for names of IT specialists who could help him.

Eventually, Kellerhals came up with a list of eight IT consultants, ranging from Big Five accounting firms to smaller local outfits. He then held what amounted to an IT "beauty contest." In a two-week period in the spring of 1997, "contestants" paraded through his office to showcase their talent.

By the end of the technopageant, Kellerhals realized that the upgrade was going to cost him a lot more than he had expected. "Their price tags made me gulp," he recalls. Some IT teams quoted bids as high as $500,000. Kellerhals, who had incurred significant debt to buy the company, thought the project would cost him about $50,000 but was prepared to spend as much as $100,000.

He needed to decide whether he wanted to build or to buy: should he have a consultant build a customized system for him, or should he buy programs off the shelf and have someone install and modify them? There were pros and cons to each approach. A customized system would have the obvious advantage of addressing the company's specific needs, but Kellerhals was concerned about cost and the duration of the project. Buying software off the shelf would save money and have the advantage of the programs' having been tested and proven. But Kellerhals needed software that could handle retailing, manufacturing, and catalog sales. And because he made a food product that required perennially fresh inventory, much of the available inventory-tracking software was inappropriate. "There is no 'chocolate software," Kellerhals concluded.

Bryan Sapot, consultant: "He [Kellerhals] had a clear vision of what he wanted. He just wasn't quite sure how to get there."

In the end, what Kellerhals took away from his search for IT help was a good idea of what he did not want to do: change the way the company did business in order to meet the requirements of new software. "When I came into the business, I was asking for a lot of change from the employees" in order to ramp up growth, he says. "I was a little afraid of bringing on too much change." He didn't want his candy makers, who were not computer savvy, to have to learn too many new things on the computer. The costs of retraining his staff -- both psychological and financial -- were prohibitive. "We could've built a rocket ship here but then had no one to fly it," says Kellerhals. "I didn't want the new system to have a big impact on the end users."

Kellerhals's head was spinning. His office manager, Kortney Goodman, who was compiling notes from the IT tryouts, approached the CEO with a suggestion. "She told me, 'My fiancÉ does this [kind of work]. You might want to talk to him," Kellerhals recalls. He knew that Bryan Sapot, Goodman's fiancÉ, was young (he was 23) and relatively inexperienced. But Kellerhals thought that, at the very least, it might be helpful to use Sapot as a sounding board for the various proposals that were before him.

Then Sapot came in to see him. The young man was looking to go out on his own and decided to make a pitch for Kellerhals's business. He had a near home-field advantage when it came to Bissinger's problems. "My girlfriend would come home at night complaining," Sapot says. "The order-entry system would crash all the time, especially around holiday time. She'd have to stay till 9 or 10 at night dealing with it."

Rather than buying new materials or building from scratch, Sapot suggested, the company should take its three systems and develop a way for them to talk to one another. "My background was in integrating systems," Sapot explains. "You don't have to replace anything, you minimize retraining for the users, it's more cost-effective, and it's an easier transition. You're fitting the software to his business model rather than the other way around."

Sapot's homegrown approach resonated with Kellerhals because it involved the least amount of change for his staff. "That's why he won the day," says Kellerhals. Sapot's price was right, too: he put in a bid for $58,000, which he now acknowledges was an attempt to undercut the quotes of more experienced competitors. And since Sapot's plan was to build on the existing Unix-based system rather than use the Windows-based systems that most of the other consultants had recommended, Kellerhals did not have to invest immediately in new PCs.

"People said to me, 'You need to have your head examined, trusting this guy," Kellerhals recalls. He admits that going with an untried consultant wasn't the safest decision. "But I knew this was right for me. One big factor is trust. He was someone I felt comfortable with."

Still, given Sapot's relative inexperience, Kellerhals did take some precautions. He drew up a contract that said Sapot would be paid upon completion of each stage of the project, the parameters of which were spelled out in great detail. The contract specified an end date and monetary penalties that Sapot would incur if the project wasn't completed on time. At one point, after Kellerhals and Sapot had become comfortable with each other, they switched to a monthly-retainer system. Kellerhals now calls that "a big mistake." Without specific deadlines built in, the project veered off schedule. They quickly switched back to a deadline-oriented pay schedule.

Sapot's tweaking, upgrading, and creative preservation have left Kellerhals with the tools he was looking for. "We can make these extensive, beautiful queries," he says. "We have the ability to pull information out of the system" that helps with planning workers' schedules, maximizing profit margins, and tracking costs. He can also examine the sales differential between a March Easter and an April Easter. "I didn't even know these pools of information existed," he says.

Information is in real time, so Kellerhals can click on a Shipping Costs icon and figure out how much he's spending on shipping on any given day. Under Sales Information, he can find out instantly how many boxes of pecan nut balls he's sold versus how many he sold on the same day the year before. He now bases his production schedules on that information, and because the system can flag items that are selling at a fast rate, he can produce more of what's needed, thus avoiding shortfalls.

Understanding the costs and profitability of each product allows Kellerhals to make informed decisions about whether to stop making certain candies. "We made the best peanut brittle in the world," he says. "But we stopped making it because the costs were too high. We were not making enough of a margin on it." But with the company's highly popular molasses puffs, which take three days to make, he continued the product and raised its price because he felt customers would be willing to pay more for it. "Before we had the computers, we never would've made those decisions. We would have just gone by our gut," he says.

The biggest savings, according to Kellerhals, have been in scheduling labor and use of the plant to maximize producti-vity. He realized, for example, that during the months that are typically the slowest for candy makers -- the summer and early fall -- he could use the almost idle plant to make seasonal products dependent on fresh fruit, such as chocolate-covered raspberries and blackberries, which are now big sellers for the company.

Kellerhals has been eminently pleased with Sapot's work and continues to retain the consultant and his new company, Datix, for all his tech support. (Kortney Goodman, Kellerhals's former office manager, is now Kortney Sapot, Sapot's wife and his project manager at Datix.)

To date Kellerhals has spent a total of $123,000 for his technology upgrade, including some Y2K compliance and the company's new Web site (, and he feels the payoff for his company has been huge. Sales were up to $6.5 million last year. By minimizing overtime, human errors, and production costs, the technological upgrade added about $500,000 to Kellerhals's bottom line and three to four points to his overall margins. The overhaul was a huge eye-opener. "It was really about taking us out of the Dark Ages," he says.

Please e-mail your comments to