Andre Martin (right) and Darryl Landvater of RedPrairie Collaborative Flowcasting Group
By André Martin and Darryl Landvater
Distribution resource planning (DRP) has been helping consumer packaged goods companies and other manufacturers with supply chain planning for decades. Every major manufacturer is using DRP, and the premiere supply chain planning vendors embed DRP into their products. DRP is the de facto industry standard.
Now the promise of DRP has reached the storefront, extending its capabilities across the entire supply chain. Called Flowcasting, it applies DRP concepts to the store and facilitates collaboration among manufacturers and retailers. The technology uploads retail point-of-sale (POS) data on a nightly basis, creating granular, store/item-level demand forecasts, shipment plans and projected inventory levels. Flowcasting aggregates those shipment plans into retail distribution demands using a dependent demand calculation.
The same process is applied to the retail distribution center, which in turn creates the demand on the next level in the supply chain. The concept of dependent demand is new to retailers but has been tested and proven in manufacturing for over 30 years and is central to both DRP and manufacturing resource planning (MRP) systems. Basically, a dependent demand is the mechanism for integrating the supply chain.
Rather than forecasting at each level (disconnected), the planned shipments at one level become the demand at the next level (an integrated supply chain). Not only is this more accurate, but it also eliminates the "bullwhip" effect and provides nearly instantaneous visibility up and down the supply chain. If a product sells more than expected at the stores today, manufacturers can see whether or not their existing production schedules will support their customers by the next morning. The data reside in one system, leveraged by both supplier and retailer.
Not only do these plans create an order of magnitude greater accuracy throughout the multi-enterprise supply chain, but for the first time, manufacturers and retailers are looking at the same supply chain.
From our perspective, Flowcasting will eventually replace the reorder point/execution systems that retailers rely on today and that manufacturers used some 30 years ago before planning systems like DRP appeared. What took retailers so long to adopt Flowcasting? Well, the sheer processing power to accommodate retailers' needs was just not available until recently. Data volumes in manufacturing are generally much smaller than they are for retail, where the number of product/location combinations can reach into the hundreds of millions. Traditional DRP systems were never built to handle these types of store-level data volumes. As a result, trying to take a platform designed for a distribution center (DC) and using it for the massive data volumes at stores was unrealistic.
With Flowcasting, trading partners can aspire to fully integrate their supply chains from their retail stores all the way back to the factories. Moreover, this type of technology can handle the required several hundred million products/locations to allow planning a year into the future, both on today's hardware and at an economical price.
The vision for Flowcasting has been around for a long time. But interest is just now growing — not only because of the technological breakthroughs, but also because of its demonstrable store-level capabilities and business results.
Driving Improved Forecasting
Forecasting accuracy is one of the most difficult things to bank on and one of the major drivers of cost in the supply chain. Among the largest costs to manufacturers on a day-to-day basis are the surprises coming from the demand side. After all, manufacturers are the ones who have to cover the uncertainty in demand with safety stock. They face the challenges of changing their production schedules on short notice as they confront unexpected demands. These scenarios affect not only the productivity of manufacturing operations, but also the productivity of other parts of the organization, as people scramble to make the changes necessary to satisfy retailers. Bottom line: These changes add costs to the supply chain.
Ralph Drayer of Proctor & Gamble, a significant contributor to improved supply chains over the years, said that if his company could stabilize its master schedule — not necessarily solidify it, but just stabilize it — then it would see a 30 percent improvement in productivity.
That's just what is happening with Flowcasting, which has enabled a large CPG company and a major retailer to integrate their supply chains. This specific CPG/retailer supply chain has four levels: more than 500 stores, 40 store-directed DCs, and 30 manufacturing-directed DCs supported by dozens of factories throughout the U.S. and Canada.
With a single view of the supply chain, the manufacturer/retailer alliance was recently able to predict retailer orders on the manufacturers six weeks in advance with an absolute error (+ and - do not cancel out) of +/- 2.3 days for three consecutive months. That's music to the ears of anyone who has responsibility to schedule production. If you can tell people in production what they have to make with that degree of precision, they won't believe you at first, but they will certainly be intrigued. Whatever their views, the numbers don't lie.
As a result, both the CPG organization and retailer desire expansion of their Flowcasting experiment. The CPG concern is expanding into another large retail account so it can achieve better global inventory optimization. The retailer, meanwhile, was so pleased with the results from Flowcasting that it would like to see other suppliers do the same thing.
Fostering Visibility, Stronger Relationships
Flowcasting is all about being able to plan the entire supply chain inside a single software system. If one looks at the cost of managing a retail supply chain, the single largest variable is the cost of producing a product.
From all perspectives, Flowcasting delivers a better forecast. Manufacturing executives can look at it, and when they're off course, they can pinpoint the problem. Was it a change in order policy, a change in forecast, or a forecast blown out of the water that forced them off-course? They can find the root causes to problems and fix them.
But that's not all. A manufacturer benefits from healthier relationships with its customers, too.
Here's one example. If retailers wanted to increase display quantities in the past, then they would submit a big order. This creates two problems: shipping the order on time and preventing the forecasting system from overreacting. Often, manufacturers see a spike in demand and increase production accordingly, when in fact no change in demand had occurred. In Flowcasting, the spike in orders can be accurately simulated ahead of time, and the manufacturer can reliably assess whether or not it can meet this increased demand. If not, then it will work with the retailer to create a realistic plan. Since Flowcasting is a model of the business, there are no "adverse side-effects" to be managed.
So, a great side benefit of adopting Flowcasting is that conversations between manufacturing and retailing executives are taking place where they weren't previously. The effects of changes at any level in the supply chain are known in a few minutes. There's a clear picture of what's going to happen. In the past, these types of issues hit manufacturers like a ton of bricks. Today, with Flowcasting, an accurate picture is emerging showing what those changes in the supply chain mean. If manufacturers can execute those changes, then great; if they can't, then they open up a discussion of alternatives with their retail customers.
In the past, manufacturers and retailers had their own supply chain systems that managed their own DCs and factories, and those systems didn't talk to each other. So there was no visibility on either side of that supply chain. Flowcasting eliminates this conundrum and gives trading partners visibility.
At the store level, forecast accuracy improves by evaluating the causes of errors (problems with profiles, assumptions made with promotions, competitive products, effects of the economy, etc.). Still, retailers may see little reduction in inventory carrying costs because inventory levels at the store level have less to do with what is needed statistically, and more to do with what looks good on a shelf. At the DC level, however, a 30 percent reduction in inventory is possible — and on the manufacturing side, a significant reduction in safety stock levels and improvements in manufacturing productivity.
Because Flowcasting provides a manufacturer with a forecast at the store level, performing the inventory calculation for the DC level is much more accurate. In our earlier example, the service level at the DC also saw substantial improvements, with a more than 70 percent reduction in out-of-stocks, improved inventory turns and a substantial increase in year-over-year sales.
All in all, a combination of productivity improvements, reduction in safety-stock levels and a stabilized master schedule alters relationships for the better. It's a win-win situation for all players.
Besides these points, if a retailer has confidence that the manufacturer can produce on time and can deliver consistently to the plans, then the retailer is more likely to tap that supplier if it has a problem with a promotion or other issue. That's competitive advantage.
Years ago, we surveyed CEOs using MRP and DRP systems who indicated that the primary reasons they went with such technology was for improved control of their businesses. To justify the expense, they touted the benefits of improved service levels, inventory and productivity. As it turns out, these perspectives are consistent with what is transpiring now at the retail level.
Elevating Service Levels
Improving service levels on the shelf translates to substantial improvement for the average retailer. If a retailer is running 85 percent in-stock, for example, and can take that number up to 95 percent, assuming a 50 percent substitution rate, then it has increased sales by something on the order of 5 percent. The result is that all the gross margin goes straight to the bottom line because the retailer already has paid for expenses relating to people, facilities, trucks and all those fixed costs. Moreover, the retailer has higher in-stock levels at DCs, in-stock service levels at stores, reduced safety stock, much better execution and a better ability to make plans happen.
People who develop good plans see better execution, too. Flowcasting gives retailers a planning horizon a year into the future — well beyond the order cycle. And that's another competitive advantage.
One way to think of a plan is as a forecast of activities. How much inventory does a retailer need to position in its stores over time? How much replenishment needs to take place at every level of the supply chain? How much does manufacturing need to align with all this to keep everything synchronized? These are the questions that Flowcasting addresses.
That's not to say there aren't challenges to widespread adoption of this technology. The No. 1 challenge is, and will always be, people. Will professionals change their behaviors and be amenable to new business processes? Other challenges include the accuracy of the store data and the availability of computers with the requisite processing power discussed earlier. (By the way, cloud computing is a viable antidote to this latter issue.)
In spite of these challenges, however, Flowcasting is changing the way that retailers and manufacturers do business with one another. While the supply chain will always be a complex animal with which to grapple, Flowcasting empowers people with dynamic, accurate information to make decisions, enabling greater collaboration between trading partners and greater overall business success. ¦
Flowcasting empowers people with dynamic, accurate information to make decisions, enabling greater collaboration between trading partners and greater overall business success.
About the Authors: André Martin (firstname.lastname@example.org) is CEO and Darryl Landvater (email@example.com) is the co-founder of the RedPrairie Collaborative Flowcasting Group, a joint venture with RedPrairie, which delivers productivity solutions to help companies around the world with inventory, transportation and workforce management.