Timing is everything it seems. It's true in buying stocks. In launching new products. It's usually the case in sports. And in comedy, of course, many a good joke has fallen flat because of poor timing. Timing also may be of great importance in software and solution investment. It certainly is for Sunsweet, the manufacturer and distributor of prunes and other fruits and fruit juices.

A year and half ago the Yuba City, Calif.-based company began looking into buying an integrated suite of supply-chain planning software. Its intention was to have everything online by this November. Sunsweet and its provider, Supply Chain Consultants of Wilmington, Del., say they are right on schedule with the implementation. Several modules of SCC's Zemeter suite already are up and running, coinciding perfectly with the picking and processing of this year's plum crop.

The impetus for the investment was cost reduction, but that was at a time when Sunsweet's primary ingredient was in abundant supply. Cutting costs is of even greater concern now. Whereas the industry has seen one bumper crop after another for several years, volume is down significantly this year. That has implications up and down the Sunsweet supply chain. Such areas as order fulfillment, procurement, packaging, inventory planning, distribution, and of course, marketing and advertising, are affected. A solution designed to optimize demand planning, and inventory and supply planning, as well as daily plant scheduling, couldn't have come at a better time.

"Sometimes it's better to be lucky than good," says Harold Upton, Sunsweet's vice president of strategic business processes. He says no one foresaw this year's small crop size, or that Zemeter would get such a workout so early on. "But since we've got that tool, we're going to leverage it, and it's really going to be a test as to what it can do."

What Zemeter already has done, at just past the halfway mark in the implementation process, is give Sunsweet's planners much more information so they can build demand-forecast reports. It's issuing inventory level alerts. Most important, as production lines have been cut, there are fewer machine, packaging and other changeovers on the lines. Production overruns have dropped from 30 percent to about 12 percent, Upton says, with 10 percent being targeted in 2005.

It wasn't always this way. Until Supply Chain Consultants was brought in, Sunsweet, a grower's-owned cooperative that's been around since 1917, handled its planning and forecasting the old-fashioned way, manually with Excel spreadsheets. And it was pretty good at it, according to Quinn Freeman, senior consultant at SCC. "They had a reasonably good process and a good idea of where they wanted to go with things," Freeman says. "But they didn't have the technology tool to do those things. What we did was take some of the manual decision making out of their hands and let Zemeter help them make those decisions and guide them in what would be the right products to make at the right time, in the right amounts to make."

Supply & Demand
It's important to remember that while a cooperative like Sunsweet is similar to other consumer packaged goods companies, it differs in a crucial aspect: it must take its "ingredients" from its grower-owners. So while its forecast drives its planning cycle in production and inventory, it can't quite match supply with demand the way, say, a shampoo manufacturer can, says Tom Garland, director, supply chain. A hair care product manufacturer buys just the right amount of ingredients based on a forecast.

"That can be our largest issue," says Garland. "When we have bumper crops, we must manage the sales through the year to utilize the crop and come out with a reasonable size of inventory. Conversely, when we have small crop year, how do we manage with a crop availability that might not necessarily meet all our sales demand?"

Sunsweet also works differently than non-cooperative, prune packagers, who can meet their demand by buying from independent growers on the open market. Of course, that is what Sunsweet does with the apricots, apples and other non-plum fruits it uses in many of its 300 SKUs.

Procurement aside, Sunsweet's manufacturing and distribution network is very much like that of a typical CPG company. It has two plants, several co-packers, and warehouses spread from the West Coast to Texas to New Jersey to Chicago. With international sales accounting for about 30 percent of its business, it has extensive distribution operation from the ports of Oakland and Houston for shipments to Asia and Europe.

"Even though the size of our company is relatively small," says Upton, "we are considered mid-sized on our sales side, yet we have all the complexities of the big guys: We do worldwide distribution, we have a lot of different channels, we have multiple plants, we have multiple SKUs - all in a cooperative environment like we have here where our supply is not our choice."

The company's needs then were fairly clear. The solution suite needed finite planning capability - which Sunsweet defined as daily plant scheduling - and inventory and production planning and sales forecasting.

Before Zemeter was implemented, Sunsweet converted raw data from its SAP backbone into forecasts and production and distribution plans.

Building and maintaining the databases was an effort, but error-ridden data entry wasn't a real consideration. The main driver for change was the slow nature of the process and its limitations on 'what-if' analysis, according to Bryce Treese, manager of inventory/planning. Sunsweet needed the flexibility of an advanced planning tool that could handle the equations required to optimize the network based on constraints it has in its plants and logistics operations.

Under the old scheme, taking on a new customer could be problematic, Garland says, because it couldn't be quickly and efficiently determined if its demand could be met. "We wanted a system that could analyze and give us the best plan back when we decided to go after this extra business or launch new SKUs."

Inadequate planning tends to create costs, so pursuing what Garland calls the "next level of sales" wasn't always the money maker it could have been. "We would try and squeeze something in so we could make the sale, but it was creating added costs along the manufacturing side and the logistics side."

Seeking Solutions
Sunsweet managers say they knew what they needed. No third-parties were brought in to advise or guide them through the software selection process. The company as a whole had already gone through an SAP enterprise resource planning system implementation, so the staff was familiar with solution deployment.

The parameters were these: A fully integrated package of functionality was needed; it had to mesh with that SAP backbone; the system had to have a good CPG focus; and cost return on the investment within a year was highly desirable.

The search got under way around June 2003, with acquisition set for the beginning of 2004 and final deployment of all modules no later than December 2004. Upton says Sunsweet wanted to see some benefits this fiscal year; again, that was before it became apparent that the plum crop would be off.

The company looked at six or more packages before opting for SCC's Zemeter product, and the obvious question is why Sunsweet didn't utilize planning functionality it already owned as part of its SAP licensing agreement. The bias, Upton says, was to go with SAP unless there was a compelling reason to do otherwise. Speed and ease of use ultimately were the deciding factors in favor of Zemeter.

"Finding the right size of tool is really key for long-term success," Upton says. "Even though we have a very complex environment in terms of what we have to support we have a very small staff. That's probably why Zemeter did so well with us. We seemed to fall within the niche market that they are after, and that's really a smaller group that has to deal with complexities. SAP has tremendous capacity and functionality, but it's a little more difficult to administer the architecture in a small environment. It became apparent that Zemeter was faster to implement - definitely a lot less time maintaining the system and a lot more time using it."

Deployment
Sunsweet's five-phase rollout includes:
1) Demand visibility, which includes defining the hierarchy of information and attributes - ship-to, by sales office, by channel, etc. It also involves creating the interfaces to move the information from SAP into Zemeter.

2) Demand planning functionality: creating statistical models, determining analysis of sales forecast accuracy, etc.

3) Inventory planning: inventory data captured by day, by warehouse, by channel, etc. It's also where safety stock calculation is performed.

4) Supply planning module (the phase under way in late September): This module will perform the weekly and monthly production plans.

5) Finite scheduling module: the daily line scheduling module.
As a cost saving measure, SCC isn't on-site at the Yuba City headquarters full time. Rather, its people are doing off-site modeling and then generally come to California for two or three days every other week.

The training that's involved largely is one of weaning the staff from manual, paper-based ways of doing things. That's no small thing because they have to understand the new concepts and software that SCC is bringing. Ultimately, they will have to take responsibility for a specific module.

Zemeter has to be integrated with SAP, of course, but that mostly has to do with data exchange rather than functionality. "It's really a rules-based type of implementation," says Upton. "There is some report writing you have to do, but you really aren't designing or configuring this implementation as you would with SAP or Lawson."

For its part, SCC is aware that one of its challenges is convincing users, not just at Sunsweet but elsewhere, that they don't need to go back to spreadsheets.

"We've learned in supply-chain planning that the technology has to be much more useable than the first-generation products out there were," says Jim Heatherington, director of business development for SCC. "What we've tried to do is create a product suite where the user community says, 'I can do this.' I think that's what a lot of directors and VP's of supply chain are looking for: 'Will my people use the tool or are they going to migrate back to Excel?'"

Wearing his salesman's hat, he says two attractive things about Zemeter (and attractive to Sunsweet, in his estimation) are its "sustainability" and its basis in the Microsoft platform.

Heatherington says the product has a lot of "what I want to call configurability" built in. Accordingly, power to change the way the tool is used is pushed back to the user, and that provides a sustained interest in the product. "It gives a lot of independence to the user, and that's what I think they are looking for."

Zemeter can be changed much faster to meet the needs of a given business process because software doesn't have to be recoded, Heatherington says. "So with Sunsweet it wasn't a matter of redefining all of their processes. I'm not going to pick on SAP, but quite often with products that are more inflexible in how they are implemented, there's a standard process: People are going to have to change the way they plan or you're going to have to change the software code. This is a big differentiator for Zemeter."

Technology-Ready
The Microsoft platform - the relational database is SQL server - means that IT groups don't have to learn a new language.

As Garland and Upton note, these things clearly were attractive to Sunsweet and largely tipped the decision in favor of SCC. But Heatherington acknowledges that Sunsweet had canvassed the competition as well and already had a good idea of whose products matched their needs.

"I give these guys a lot of credit, because they were ready for advanced technology," he says. "When you come in on any project, there is always readjusting and trying to align user requirements. But they weren't at square one when we walked in the door. Internally, they had a very strong group that knew what it wanted."

From the standpoint of SCC, things have gone quite well. Heatherington and Quinn say they were upfront about how long the initiative would take and the likelihood of meeting that timeline. They also know that Sunsweet wanted some results as soon as possible, and SCC endeavored to make that happen.

"We try on every project to give clients a deliverable within six week," Heatherington says. "That's not necessarily rolling out a entire module. What it could be, for example, is if a company were looking for better forecasting and demand planning - they would be able to look at their data and analyze it, do the stratification and volatility analysis, and metric reporting very, very quickly.

"That's good for morale, it's non-disruptive to the company, and they know they are talking about a project that's going to take weeks, perhaps months, but certainly not a year."

SCC likes to get data, particularly on the demand planning side, in the hands of its clients quickly. Typically, here's how it works: After meeting with a client, a list of attributes is drawn up to determine how it needs to see its demand data presented. The client's ERP system is then tapped for data, which is ported to Zemeter. Within two or three weeks, the client is able to extract that data from Zemeter, though tweaking and further refining may be necessary.

"We usually start with demand data because that tends to be where a lot of the problems are," Heatherington says. "We then slowly move into more complex phases along the spectrum and that tends to mitigate those types of issues." He says the Sunsweet project has not had a "hook" to delay the phased rollout or that's been "overly cumbersome" on the client.

What's the benefit of all this, from the standpoint of SCC? Heatherington thinks it may be difficult to put a dollar figure on problems that are avoided by employing a system that gives company-wide visibility. But there is a distinct difference in a system that gives a report on inventory levels in a warehouse and one that 'trends' this data daily. "I can take that data and marry it against my orders and marry it against my forecast. That tells me that next week I'm going to have a problem here, and I can take corrective action. That's something that, frankly, supply-chain planning doesn't get enough credit for doing."

Sunsweet, faced this season with a smaller plum crop than expected, is faced with just that, taking corrective action, to keep supply and demand aligned. In a manner of speaking, Zemeter's supply-chain planning capability will get a real workout as soon as it's out of the box. Its coming on line couldn't have come at a better time.