Demand planning isn't what it used to be. It's harder. Even with the help of the latest supply chain execution tools, managers need a strong combination of market knowledge and partnership skills to anticipate and plan to meet consumer demand.
That premise was the focus of a roundtable discussion MH&L moderated among members of the Material Handling Industry's Supply Chain Execution Systems and Technologies Group. This meeting took place at MHI's annual Fall meeting and participants identified key trends managers need to watch if they are to keep up with the whims of online consumers.
Discussion participants were:
Jim LeTart, director of marketing for RedPrairie;
John Reichert, director of product marketing for Tecsys;
Dinesh Dongre, vice president of strategy with Softeon; and
Alan Reigart, vice president with St. Onge.
Serving Multi-Channel Demands
Jim Letart started the discussion by identifying the various sources of demand.
"We went from defined channels where demand planning was based on what was happening in a channel to now an environment where you have many channels and it's not one demand anymore," he said. "There's a whole different series of demands and the fulfillment is now mixed. You're no longer in separate channels and you can't plan each one of them separately.
"Demand planning has to take all the information in from the different sources, some of them new, like mobile," he added. "We also have to fulfill from many sources yet keep all those numbers intact so you can have a demand plan that makes some sense. Mixing those channels together as if they are one is the biggest challenge retailers face today. And they don't know how to do it right now."
The challenge is, multi-channel fulfillment includes direct-to-consumer as well as third-party-managed. Traditional demand planning systems weren't designed to factor-in these channels. They were designed to plot demand patterns based on what one's own DC shipped.
ohn Reichert said it's important to interpret the nature of these demand shifts so that future business can be plotted.
"As demand shifts from one channel to another you have to recognize whether you are adding new volume or if demand is just shifting from one channel to another," he said. "That way you can start to analyze not only where it's coming from or where it needs to go but what is the source. When you're looking to expand your market this will tell you if you're actually capturing a new market."
He predicts that more managers will be able to analyze the curve from a more traditional plan to an e-fulfillment plan over a certain time span and then start to change their physical distribution network to match that forecast.
Retail as Role Model
As such demand planning best practices take hold in the retail world they will set the example for other industries to follow, Reichert added.
"Healthcare providers are so far behind on the whole supply chain that they're starting to adopt retailer type practices," he said. "They're starting with the idea that they don't have point-of-sale data at the point-of-use. They're starting to expand their networks very much like e-fulfillment where you have distribution to a hospital and to a clinic. You may go to home care, and orders may get distributed similar to an e-fulfillment model, from any point in the chain. They don't have any of that demand sensing at the front line, so some of the demand planning systems and applications being built on the advance side will really help healthcare."
As sophisticated as retail demand planning is becoming, however, Dinesh Dongre says historical demand isn't as relevant today because of demand's volatility. Planning must take into account the short term as well as the long term. He identifies three components of planning: planning, responding to feedback and optimizing.
"Networks are becoming more global, with outsourcing not only of suppliers, but of manufacturing and even fulfillment," he said. "That means a lot more nodes need to be considered when looking at an order-to-cash and order-to-fulfillment cycle. With demand variability the product segmentation complexity also comes into play. All of these things are driving the need for critical and robust planning for near and longer term planning."
The Human Element
While it's important to plan for what happens outside the distribution center, the variables inside it must also be considered, suggested Alan Reigart. Specifically, factor in the labor element.
"Sometimes we overlook the professionalism of a fulfillment operator in a warehouse," he said. "What do you expect to get done as well as when and how well? The key is on-time and the second is order accuracy—achieving the same levels you have today while changing to new models."
However, the pace and complexity of multi-channel supply chains are outpacing the ability of human beings to absorb them. That's why system vendors are building in intelligence so the buyers and the people making the shipping decisions don't have to deal with those decisions. Instead, they get pulled out of the system.
"From an application standpoint we have to bring the sophistication so the person trying to make a decision gets the end result on what they need to buy and where they need to push it," said Reichert.
That requires visibility, added LeTart.
"Because you have all these different channels and all these input sources, you have to be able to see that demand from all the different sources and it has to be near real time," he said. "The volatility is so high you can't take the traditional batch approach because it's not fast enough for today's changing environment."
Daily Adjustments and Shaping
Alan Reigart agreed with LeTart's assessment, adding that as things change, planners need to make sure they're not cannibalizing what was planned for another channel. Beyond such visibility today's demand planning solutions need a system on the back end that can handle the volume of information that makes visibility possible. That's more than demand planning had to do before.
"You have to look at all that movement, both the fast and slow movers, and be able to predict, based on that flood of information, what your demand pattern will look like going forward," Reigart said. "This is so you can predict it and readjust every day."
Dongre concluded that demand planning needs to advance from historical data analysis to "event shaping." That means rather than reacting to demand changes, actually driving out demand changes.
Demand sensing and shaping, and in a broader sense, harmonizing plans across the supply chain network, will enable predictive analytics and give planners a more solid base on which to build their supply and demand strategies.