DDMRP (Demand Driven Material Requirements Planning) is a term popularized and promoted by the Demand Driven Institute (DDI). Chad Smith and Carol Ptak, co-founders of the DDI, have written several books on the topic of DDMRP, including Orlicky’s Material Requirements Planning (Third Edition), DDMRP – Demand Driven Requirements Planning, and their most recent work, Precisely Wrong – Why Conventional Planning Fails and How to Fix It.
In a recent conversation with Mark Robinson, one of the co-founders of Orchestr8, he expressed both his high regard for the Demand Driven Institute, and his concerns about how the demand-driven thinking can move forward from where it is today.
The vast majority of successful and published DDMRP case studies are relatively small in scale. Generally speaking, they are confined to impressive improvements within a factory or across a small part of a supply chain network. What we are not seeing case stories about enterprise-level change.
What is the problem?
Robinson admits, “I am very concerned when we speak to potential DDMRP implementers, particularly about their understanding of the software tools required to facilitate an enterprise-level rollout.” It is not sufficient to simply “calculate some buffers, and bolt these onto your MRP system.”
“We were” calculating demand-driven buffers “back in 1997” [20 years ago] and bolting those calculations onto traditional MRP and ERP systems, and doing it “quite successfully within a single factory,” Robinson told us. The problem, however, according to Robinson, is that doing so “doesn’t scale to the enterprise level.”
How and why can Robinson say this?
Well, let us begin with the fact that Mark Robinson has more than 20 years of supply chain management experience, having held positions with Lucas Engineering and Management Services, Computer Sciences Corporation  and World Class International [WCI Group]. Then, of course, there is the fact that he co-wrote the winning entry for the 2002 Supply Chain Council  Award for Technology Excellence.
The problems, Robinson believes, are on several fronts whenever demand-driven is attempted at the enterprise level.
Robinson is convinced that there are at least five crucial reasons many DDMRP-compliant solutions would fall down when applied in an extended enterprise.
Only one planning tool can be allowed. The idea of a bolt-on system, with planners dipping in and out of both systems simply does not work at enterprise level. Therefore, the legacy system, whether SAP or another, whilst still acting as the backbone system and master source of data, must no longer be used by planners for order management and eexecution. The implication of this is that whatever DDMRP tool is used must then provide the breadth and depth of functionality to facilitate this change across both demand- and supply-side planning. While dependencies upon the legacy planning systems can be reduced gradually with the introduction of DDMRP, the ultimate aim must be to end such dependencies. Actually, most ERP planning platforms are not planner friendly, or particularly cost effective, so the change may not be a difficult one to contemplate.
Experts have long acknowledged the inherent “system nervousness” engendered by MRP’s usage both for planning and to drive execution. Because MRP attempts to create end-to-end dependencies through the bills of materials (BOMs) and schedules (the MPS to which the MRP system is directed), relatively tiny changes in the underlying data (e.g., forecast demand, actual demand, resource availability) may lead to large swings in calls for action—even contradictory actions.
Most supply chain managers and planners in enterprise-level organizations are required to supply some kind of long-term plan. Therefore, it is entirely appropriate to have a mid- to long-term supply plan generated that uses the same logic as the Demand Driven replenishment signals. The results is that everybody on the supply chain sees a consistent view of capacity and supply, rather than the problem of a step change in supply at the lead-time horizon, which we see frequently experience when firmed DD orders are wedded with MRP planned orders.
At the enterprise level, additional crucial inventory planning rules must be accommodated—beyond the buffers provided by typical DDMRP-compliancy models. Robinson includes 1) demand-driven reorder cycle, 2) rate-based planning, 3) forecast driven planning, and 3) spares planning (among others) in his list of those typically missing from many DD-compliant offerings. “Without a spectrum of planning rules to apply, product segmentation using volume and variability is pointless,” Robinson suggests.
The level of functionality and general logic around supply planning in ERP systems, were conceived in the 1960’s. As a result, key business processes are handled very poorly relative to today’s supply chain world. For instance, life cycle management, event management, ordering groups, production wheels, transport load building, etc. are all very limited or missing entirely in typical ERP systems of today. All of these, however, represent day-to-day challenges for most enterprise-level planners. If your DD planning tool does not have the scope to cover these key areas, you will end up trying to shoehorn the DD processes into the ERP tools, which are already proven to be not process capable. The result is vastly diluted effectiveness. This gives rise to the business case for extended enterprise-capable demand-driven solutions.
“Any or all [of the above]” might lead to “diluted benefits from the expected areas of inventory reduction and [improved product] availability,” Robinson adds. “But more importantly,” he says, in the absence of such capabilities, a move to demand-driven may actually “jeopardize the likelihood of success of the project by [actually] increasing planner workload and process complexity.” This would be just the opposite of expectations and, naturally, could damage the reputation of the entire demand-driven theory and practice.
Mark Robinson wrapped up on conversation by agreeing that there was much more to be covered on this topic demand-driven at the enterprise level. So, be watching for future articles addressing these important concepts and issues.
Please feel free to leave your comments here.
 Now Computer Sciences Corporation is now DXC Technology after a merger with the Enterprise Services division of Hewlett Packard Enterprise.
 The Supply Chain Council (SCC), originally independent, has been incorporated now into APICS as the APICS SCC.