Demand Planning at Coca-Cola: What's the Secret Formula?
The sign in the elevator banks at Coca-Cola Refreshments reads: "There is a person at the end of our supply chain." Well, that's a start.
If only it were that simple to sell 1.7 billionservings of sugared water a day. But, like just about every big corporation with a complex, global supply chain, Coca-Cola battles constantly to align its various disciplines and their diverging priorities.
"Plants want to run as much as possible. Sales wants to sell more. Marketing assumes huge successes. Logistics and procurement are on their own. Finance is saying no to everything," said Rob Haddock, group director of planning with Coca-Cola Refreshments (CCR). Sound familiar?
To be sure, The Coca-Cola Company has sought to streamline and simplify a supply chain that historically was fragmented by its very nature, with production and bottling carried out by separate entities. CCR was formed in 2010, when Coca-Cola acquired the North American business of Coca-Cola Enterprises, its largest bottler. The sales and operations side was branded as CCR, which also took in the majority of The Coca-Cola Company's U.S. and Canadian businesses.
In the process, the company began paring back the number of bottling operations, both owned and independent. The 353 ownerships that existed at the time of CCR’s formation have been gradually reduced to 100 plants and 76 ownerships, Haddock told the San Francisco Roundtable of the Council of Supply Chain Management Professionals, at a recent meeting of the group in the Bay Area. (There were more than 1,000 bottling plants in 1960, Haddock noted. The formation of Coca-Cola Enterprises in 1982 marked the beginning of the company's efforts to get that number down.)
Today, the Coca-Cola production and distribution network consists of 742 facilities, and more than 650 distribution centers. The company is still working to "clean up" the system so that individual units have more contiguous territory, Haddock said.
Complicating matters is the huge number of products that are sold under the banner of Coca-Cola and its subsidiaries. "Coke" itself comes in multiple flavors, with and without caffeine and sugar. But the entire line comprises some 3,500 products. And the company is hardly immune to the trend of SKU proliferation. "It's been total chaos since the introduction of Diet Coke in 1989," Haddock joked. Or was it a joke?
It begins, of course, with planning, and that's where Haddock and its colleagues come in. "My people are telling all the assets what to do – when to make it, how to move it, when to buy it," he said.
Their guiding principle is the need for integrating demand and supply, as summed up by the acronym DOIP, for demand, operations and inventory planning. Demand involves daily interaction with marketing and sales. Operations focuses on the financial implications of optimizing production and distribution. Inventory planning assesses the impact of striking the right inventory levels. All three areas are synchronized through the company's sales and operations planning (S&OP) effort, the forum for making and sharing key decisions.
None of this had led to a completely reliable forecast, as even the most successful demand planner will tell you. "World-class" accuracy these days is around 70 percent, Haddock said, meaning the company has to absorb that remaining 30 percent, usually by building inventory.
The actual number can be even worse than that. Take that 70-percent grade for forecast accuracy. Now assume that production will meet requirements around 85 percent of the time, and that customer-service levels will be in the range of 95 percent. Put it all together and the accuracy rate is closer to 57 percent, Haddock said.
Companies fight constantly to push that number upward. Breaking down corporate siloes, said Haddock, is a crucial first step.
When devising a forecast, it’s essential that everyone aligns to one number, he said. Still, it’s not easy “to extract the truth from a lot of people who don’t really want to divulge what the real plans are.” The minute that sales and marketing quotes a figure to senior leadership, that number “gets etched in stone somewhere up on the highest floor of the building…. So don’t blame the sales and marketing person for not being specific.”
Imperfect though it might be, the forecast has to be propagated both upstream and down. Suppliers are waiting for a demand signal, which controls master production scheduling. Then there’s outbound transportation, which operates on tight margins yet is in danger of overspending its budget when it has to scramble to fill unanticipated demand. “That’s one of our big challenges right now,” said Haddock. “How do I know that I’ve got enough trucks, day in and day out?”
Individual service-level agreements will have an impact on the forecast as well. Then there’s the issue of forecast bias, caused by a failure to adjust to changing circumstances. Inventory, for example, might be positioned to handle a one-week spike. But what if it regularly stretches to two? It could be time for a rethinking of inventory strategy.
Good forecasting and demand planning is all about combining the notion of a “perfect order” with the realities of the marketplace. Even the best plan is probably outdated by the time it’s put into practice, Haddock said. “If I’ve got 85 percent of the solution, I’ll just go with it. It’s never going to be 100 percent.”
The acknowledgment that error is inevitable is no excuse for failing to integrate the various disciplines within the organization. In addition, said Haddock, all departments must be ready to tweak the plan in line with actual events.
“We’re constantly readjusting our supply chain in response to changes in the forecast and live orders,” Haddock said.
The lesson for all manufacturers and distributors: Plan and replan. “The dynamics of the business change daily,” Haddock added, “and require subtle course corrections.”
No comments:
Post a Comment