My Take: Let's admit seven demand management mistakes of the last decade

Article by Lora Cecere from Supply Chain Shaman; published on January 28th 2013

 

Within an organization, the words “Demand Planning” stir emotions. Usually, it is not a mild reaction. Instead, it’s a series of emotions defined by wild extremes including anger, despair, disillusionment, or hopelessness. Seldom do we find a team excited about demand planning. Supply chain leaders want to improve it, but are not optimistic that they can make improvements.

After two decades of process and technology refinement, excellence in demand management still eludes supply chain teams. It is the supply chain planning application with the greatest gap between performance and satisfaction, and is the area with the greatest planned future spending. For most teams, it is a conundrum. It is a true love and hate relationship.  They want to improve demand planning, but they remain skeptical that they can ever be successful in improving the process. As shown in figure 1, demand planning is important to supply chain leaders, but also an area with very large gaps in user satisfaction.

In our research at Supply Chain Insights, we find that demand planning is the most misunderstood of any supply chain planning application. Companies are the most satisfied with warehouse and transportation management and the least satisfied with demand planning.

Teams are also confused on the process. What drives excellence in demand planning has changed and well-intentioned consultants give bad advice. In this article, we share insights on the current state and give actionable advice that teams can take to make improvements.

Why it Matters More than Ever. Facing the Supply Chain Plateau.

Supply Chain Management (SCM) concepts are now thirty-years old. The first use of the term supply chain management in the commercial sector was in 1982. Previously, the focus was on a more siloed approach to improving manufacturing, procurement or logistics. When they were lumped together, it gave birth to the concepts of demand planning and integrated supply chain planning.

The first demand planning applications were introduced late in the 1980s. Today, most supply chain professionals believe that the supply chain planning solutions have driven steady progress to reduce costs, improve inventories and speed time to market.  What we find is that we have actually moved backwards over the course of the last ten years on growth, operating margin and inventory turns. We have improved days payable, but this has pushed costs and working capital responsibility backwards in the supply chain, moving the costs to the suppliers.

To move forward, we have to admit the mistakes of the past. We need to fail forward.  In this journey to sense and shape and use demand information to drive a more profitable response, leaders have to confront a number of mistakes made in the design of demand processes over the course of the last decade. Here we start with the seven that we see the most often:

1) One-number Forecasting. It is a Hoax. :  Well-intentioned consultants tout the concept of one-number forecasting. Eager executives drink the magic elixir. But, they realize too late that this is overhyped and too simplistic. As a result, the concept adds, does not decrease, forecast error.  The reason?  It is too simplistic.  The people who push this concept do not understand demand planning.

A demand plan is hierarchical around products, time, geographies, channels, and attributes. It is a complex set of role-based time-phased data.  As a result, a one-number thought process is naïve. An effective demand plan has MANY numbers that are tied together in an effective data model for role-based planning and what-if analysis.

A one-number plan is too constraining for the organization. A forecast is a series of time-phased numbers carefully architected in a data model of products, calendars, channels and regions. The numbers within the plans have different importance to different individuals within the organization.  So, instead of one number, the focus needs to be a common planwith marketing, sales, financial and supply chain views and agreement on market assumptions. This requires the use of an advanced forecasting technology and the design of the system to visualize role-based views that can only be found in the more advanced forecasting systems.

2) Consensus Planning:  In the last ten years, the concept of consensus planning was advanced by the industry with the belief that each organization within the company could add insight to make the demand plan better. The concept is correct; but for most, the implementation was flawed. The issue is that most companies did not hold groups within the organization accountable for bias and error.  Each group within the company has a natural bias and error based on incentives, and unless the process has discipline around this reporting, the process of consensus forecasting will distort the forecast and add error despite well-intended efforts to improve the forecasting process.

I have worked with one company that has redesigned their collaborative demand planning processes three times.  Each time it was to improve the user interface to make data collection easier by sales. Not once did they ever question the value and appropriate use of the sales input or apply discipline on the input that was driving a 40% forecast over-bias. I struggle with why more teams do not apply the principles of Lean to consensus planning process through Forecast-Value Add Analysis. This is best described by Mike Gilliland in his book The Business Forecasting Deal: Exposing Myths, Eliminating Bad Practices, Providing Practical Solutions.

3) Collaborative Planning Forecasting and Replenishment (CPFR). This process was the most widely adopted in the consumer packaged goods industry. The design of the process was for manufacturers to collaborate with their retail partners on the building of a demand plan for the extended network. This process, termed Collaborative Planning Forecasting and Replenishment (CPFR), was designed to align the manufacturer’s demand plan to the retailer’s and reduce the bullwhip effect. The assumption was that the retailer’s forecast would provide better insights.

The maturity of the retailer forecast was never considered. The issue is that the majority of retailers have poor forecasts, and the process never accounted for the inherent bias and error of the retailer forecast. When a consumer product company measures forecast accuracy and holds retailers accountable for bias and error, there is usually only one retailer that measures up to the test and requirements of CPFR. This retailer is Wal-Mart. For the rest, the process of CPFR has increased demand error. Bad inputs lead to a bad output.

4) Data Model Design. Forecasting What to Make Versus Forecasting the Channel Demands. The traditional technique is to forecast what manufacturing should make. This has changed to modeling what is being sold in the channel. This difference, while it may sound trivial, is a major difference. It requires a step for demand translation. Forecasting channel demand reduces demand latency and gives the organization a more current signal. It also allows the augmentation of the forecast with demand insights to improve the quality of the forecast.  For most companies, this requires a re-implementation of the demand planning technologies.

5) Rewarding the Urgent Versus the Important. Time after time, we see companies implement demand planning technologies and improve forecasting processes, but not improve the overall results of the supply chain. The issue is the lack of training on how to “use the better forecast signal.” Most supply-centric teams are not clear.  They see it as a set of numbers to be tightly integrated; whereas, the more mature teams see it as  probability of demand to be used in their network design and supply planning models. For them, it is not as much about the specific number of demand, it is about the demand pattern and the probability of demand.

6) 80% Is Good Enough.  When it comes to a demand planning implementation, the devil is in the details. Seasonality, causal factors, usability, and the depth of predictive analytics are critical.  This can only be determined through the use of the software in conference room pilots.  Unfortunately, teams rush to implement versus spending time to understand the capabilities of the different packages. The best teams carefully evaluate the pros and cons of forecasting packages through testing in conference room pilots.

7) Focusing on “Sell-into” the Channel Versus “Sell-through.”  Most organizations are only looking at the modeling of “sell-into the channel” versus “sell-through the channel.”  By sensing demand at different channel points, and managing the inventories in the channel, manufacturers can avoid returned products and obsolescence.  I was recently speaking at the Institute of Business Forecasting (IBF) conference and a leader of a mature demand-planning process was speaking.  His comment stayed with me, “I can always get better on demand planning. We can work on this over time; however, time is of the essence to measure the velocity of product movement of every channel buffer point.  If we screw up the management of inventories and the sensing of new product launch, it is the difference between success and failure.” So many times, the concepts of demand planning are seen as passive and detached from the organization. In this case, the supply chain leader took ownership of channel demand through the channel, and has gotten promoted three times since I last heard him speak.  The shift is invaluable to the organization.

comments powered by Disqus