Part one of this series appeared in the April issue of IndustryWeek, which you can read here: Tales of an ERP Implementation: Part One
After 12 months of intense competition, testing and market comparison, Jelly Belly selected Infor's M3 ERP system to drive its business off its green screens and into the 21st century.
With that purchase, the Fairfield, Calif., gourmet candy company found itself in a good position: its market was expanding, its flavor profiles evolving and its manufacturing footprint expanding. Now, finally, it had the tool it needed to make it all work.
It had all the makings of a explosive new epoch for jelly beans.
Getting that ERP system in place, however, had all the makings of an implementation disaster.
"When we see implementation projects getting off track, it's usually because the customer has lost a clear direction of how their business really works," explains Rick Veague, chief technology officer at ERP provider, IFS North America.
"Companies coming out of these old green screen systems have lost track of how their business really works," he says. "The business is running, but the current IT system doesn't really reflect their real business processes; it just reflections portions of it around basic transactional stuff. "
The rest of the business, what really makes it run, he says, is all in tribal knowledge—field expertise picked up on the job and passed down in training, but never formalized in any referenceable system. Trying to capture that knowledge and encapsulate it into a new, updated, ERP system is the recipe for disaster, leading to expensive, time-consuming customizations designed around objectives no one clearly understands.
And that is exactly where Jelly Belly found itself headed.
"We didn't have the technology to effectively utilize our manufacturing equipment or to operate in any kind of efficient manner," recalls Dan Rosman, vice president of Information Technology at Jelly Belly. "We didn't even have a proper distribution system, so we had people try to figure out where containers needed to go and when they needed to get there on spreadsheets."
Worse, he says, no one believed the spreadsheets.
"We used to manage our whole product line-up—what we thought were making us money and which weren't—all on gut feel," he says.
So when it came time to these translate businesses processes into M3, Jelly Belly he had to first invent a formal system essentially from scratch. Which is tricky for a 145 year-old company.
However, those 145 years had provided the company with something valuable: long-serving experts who knew their jobs, knew the process and knew what they needed for the system. All Rosman had to do to program the system was to tap into that expertise and build the system around it.
But that was complicated too.
"We had a mix of people here who all wanted different things out of the system," he says. "We had some people who were gung ho and wanted to re-engineer our whole process, and we had other people that just wanted it to do what the old system did. There wasn't much room between."
There had been a strong ground swell for change at Jelly Belly growing for years. Anyone dealing with the manufacturing and supply chain side of the business—anyone involved in scheduling manufacturing cycles, raw material consumption inventory or demand—desperately needed new technology configured to help them do their jobs more efficiency.
On the other side, Rosman notes, there were some people who were perfectly happy with the old system.
"The order entry folks, accounting, even purchasing to some extent already had the functionality they need in the old system," he explains. "They had a very simple process in the old system, so going over to a GUI system with multiple screens and multiple steps wasn't a popular move."
So Rosman found himself in the middle of a fight. He had a robust tool at his disposal that was customizable enough to fit whatever the business required. But to deploy it, he first had to convince his crew to define sensible changes and accept the changes to come.
It was a crash course in change management.
"Change management is hard; anyone that tells you anything else is lying," IFS' Veague explains. "Even with the easiest software in the world, it's still hard to change—especially in the manufacturing environment."
This, he says, is where implementations can go off track.
"There comes a point when you need to rethink your business process and even how you do manufacturing," he says. "That can be difficult for some people."
On the other side, being too gung ho, he warns, leads to too much re-engineering and business revision, which can derail the whole project before it even gets started.
"You've got to be careful," he says. "If you start investing too much in custom processes and business changes, you invariably end up in analysis paralysis and nothing ever gets implemented."
The way to go, he says, is to find that sweet spot somewhere in the middle.
"I think there's a balancing act and a good path between over-engineering the new solution or making the new solution look too much like the old one because of comfort or habit," he explains. "You've got to be able to navigate that path down the middle—and of course have a clear plan and a good project governance and you'll get there."
It's a matter, he says, best solved the old fashioned way: training.
"It was a big learning curve for a lot of folks and required quite a bit of training, of course," Rosman says. "But once people got into it and saw the capabilities and configurability to it, we really started to make progress."
After the initial kickback, he says, "In the end, we didn't really have to fight either side too much, unless it was contrary to the rest of the project."
And that really was the trick, apparently.
After two years of work, having endured the selection process, the early change management fights, the implementation strategies and all of the minutiae of the process, Jelly Belly emerged an unrecognizable company.
"We used to manage the company by what we thought was making us money," Rosman says. "People would say, gee, I like that flavor, so let's keep it on the line. Maybe it cost more to make, maybe it lost money. We had no way to know. "
After the implementation, though, all of that costing and purchasing information was done through the M3 system. Even the labor was done through the system, so Rosman's team could finally see real numbers and make better decisions—he could directly affect plant floor operations and the company's bottom line with actual data, not hunches.
The first tangible results of that, he says, weren't the kind of big gains you'd expect. Rather, Jelly Belly saw some significant cuts and few major losses.
"We lost some customers through this process," he recalls. "In fact, M3 helped us lose those customers."
The system helped identify Jelly Belly's toxic customers—those customers that had been abusing the company's old, lax charge-back system to save some money at Jelly Belly's expense.
"They were the customers you want to lose—the customers that don't help you stay in business or make money," he says. "Without them, we were able to focus on customers who understand how businesses operate and could help us make money."
His team's next discover came as a bit of blow. It turns out that some flavors and products at everybody at Jelly Belly loved and assumed were profitable, actually weren’t.
"They weren't popular at all," he says. "They were losing money. So we had to either fix them or ditch them. And we actually ditched several line-ups after M3 helped us understand product cost."
And that was just the start.
"For the first time, we were able to understand our customers and their profitability and our products and their profitability," he says. "We could finally see what we were doing right and what we had to stop doing. That was a big deal to us."
Rosman ties the project directly to expanded international operations, increased profitability and efficiency—all of those positive outcomes you'd expect from a big IT makeover.
"With our old system, we would not be running this business now. We already weren't running it," he says. "It was a miracle that the business was running at all before we brought this in."