Four Reviews of DDMRP for Supply Chain Planning
In recent months four experts have weighed in on Demand Driven Material Requirements Planning (DDMRP). Here is an overview of those reviews, followed by my own short commentary, plus links to the articles at the end of the post.
- Lora Cecere of SC Insights was the first of the supply chain expert opinions I saw on DDMRP. Lora said, “The concepts of DDMRP are growing in popularity. However, despite the excitement, the current implementations are largely small and regional projects (in the original post she called them “science projects”). The software approaches and project implementations are not enterprise class. The current focus is on inside-out enterprise processes, not outside-in value networks.”She likes DDMRP flows and buffers. However, she disagrees with DDMRP’s principles of forecasting and the use of demand sensing technologies. She says she believes DDMRP “buffers are only a piece of the answer” and that “forecasting plays a vital role in determining the rules for inventory buffers.”In a separate post, she also takes issue with those consultants whom, despite evidence to the contrary, “the only path forward is to blindly deploy DDMRP.”
- Tim Payne of Gartner sees DDMRP as reflection of increased interest in what Gartner call “Respond Planning”. He says that if “companies want to get the demand/supply balance equation right for their supply chains, they will need to invest in respond planning capabilities of some sort. They will need to have a short-term planning capability that is driven by a more accurate assessment of demand (which drives toward the actual order rather than a traditional forecast).” He sees DDMRP as one possibility for filling the Respond Planning need.But he says, Respond Planning “must also work in conjunction with a tactical and strategic planning capability that ensures the right level of resiliency is built into the supply chain design and use of key resources (such as inventory).” This is what Gartner calls a strong CORE (Configure, Optimize, Respond, and Execute).Payne sees a clear need for as true a real-time demand signal as possible, but “without the appropriate configure and optimize planning activities, respond planning alone is not enough to deliver full promise. It ultimately relies on the resiliency coming out of the configure and optimize layers to help it effectively buffer execution volatility and achieve the right balance between true demand and supply at the point of execution.”
- Shaun Snapp of Brightwork – Cecere recently cited a “great blog post on the topic of DDMRP” by Shaun Snapp which she encouraged her reader to study. Snapp is very critical of DDMRP, with far more details than we can go into here.To summarize Snapp says, “Statements made by DDMRP proponents are highly inaccurate in the articles that I have analyzed with many of the statements presenting a lack of knowledge about how supply planning systems function.” After a lengthy analysis, he concludes, “Overall DDMRP has really nothing that strikes me as having made a contribution to inventory management; so why would this area be anything but more unrealized promises.”
- Stefan de Kok is affiliated with ToolsGroup, but he maintains an independent persona and blog. Lora Cecere says, “If you don’t follow Stefan, you should. He is a great thinker on inventory planning and demand management.”Stefan says that a strength of DDMRP is that it can determine where products should be stocked, but that “one of its biggest weaknesses is determining how much to stock in such locations. Safety buffer levels are determined using overly simplistic logic with arbitrary parameter values. And unless these are implemented to use either a proper statistical or probabilistic forecasts by item/location and at the appropriate time granularity (equal to or finer than replenishment frequency), they are based on nothing more than a naive forecast.”One of Stefan’s biggest concerns is that “all the boasts of value it brings are in comparison to the value of MRP. That is setting the bar so very, very low. It is like boasting to a college math major that you are better at math than a third grader.” He says that good Advanced Planning and Scheduling (APS) systems have been delivering “similar and even greater value than DDMRP has been achieved for decades compared to MRP.”
Many manufacturers are frustrated that their planning systems aren’t delivering the value they want and expect. Some of the most antiquated software, like SAP APO, are founded on planning approaches that fell out of favor more than a decade ago. So naturally those users are unsatisfied and searching for something better.
Some will be drawn to a promise that complex problems (supply chains are extremely complex) can be magically solved with simplified answers. The alternative is that there are several supply chain planning and demand forecasting vendors employing increasingly powerful technology to identify and respond to a true real-time demand signal. They bring capabilities such as stochastic (probabilistic) demand modeling, demand sensing, and machine learning to these complex problems, leveraging far greater data for enormous performance gains.
The world’s supply chains are awash in demand volatility and long tail demand – two problems DDMRP barely addresses, and especially lacks the ability to plan and guarantee service levels. Instead it offers simplified approaches that might work in the minority of cases involving fast moving, relatively consistent demand. For the rest, there are much better ways to solve these problems.
If you are interested in DDMRP, you can find the blogs described above at the links below:
- Demand Driven: Can We Sidestep Religious Arguments? by Lora Cecere, May 5, 2017
- Enable Your Technology to Support Supply Chain Planning Concepts Like DDMRP (requires subscription) by Tim Payne, October 5, 2017
- DDMRP – A Repackaging of Lean and JIT by Shaun Snapp, September 19, 2017
- DDMRP: The Good, the Bad, and the Ugly by Stefan de Kok, October 9, 2017 and DDMRP: A Fistful of Considerations by Stefan de Kok, October 19, 2017
CORE is a Gartner model which stands for Configure, Optimize, Respond and Execute. This is a planning hierarchy with a lower frequency and longer time horizon towards the top and a higher frequency and shorter time horizon towards the bottom. In the past, these layers were mostly mutually exclusive and disconnected. But they have been converging and connecting via vertical alignment.
A good example is the vertical alignment between more strategic aggregated decisions in processes like S&OP (which sits between Configure and Optimize) with operational details at the Execute level. The long-to-mid-term plan only has meaning if it translates into Execution, but there can be a disconnect—between plans modeled on aggregated demand and supply numbers and operational execution plans that need detailed data down to the daily SKU-Location level. The longer-term tactical plan may not translate well into the execution plan and the execution plan doesn’t reflect the financial plan.
This is because management collaborates and makes business level trade-offs in the aggregate. But operational plans must Optimize and Execute in a detailed model that represents real-world constraints – such as plant level production capacities and distribution centers by Ship-to SKU-Locations. Sales and finance can stay at a higher level, but for optimization and execution “the devil is in the details.”