The Growing Acceptance of Supply Planning at a Detailed Level

What This Article Covers 

• How does ToolsGroup approach planning? 
• Order batching in supply planning and the implications for detailed planning. 
• What overly aggregated supply planning means.

Planning software is changing and there is a growing acceptance of the importance of capturing the details of the supply chain.

Background

It is instructive to focus on vendors that make a particular emphasis of one approach or the other. I have written quite a few articles on the importance of moving beyond the “brand” in software selection and in companies verifying that the philosophy and bias of the software in question matches what the company wants to do. Writing on this topic brought me back to a conversation I had with Joseph Shamir, the CEO of ToolsGroup. Videos of me interviewing Joseph Shamir can be seen here.

In my discussion with Joseph, I recall him emphasizing the importance of modeling the variability of the supply chain and not simply taking averages. This issue came up when I was at a company and getting involved in analyzing the batching of the orders that came out of the supply planning engine. The client was insistent that the supply plan be in weekly buckets, however, the fact was that this was a display issue. That is the weekly buckets were just how the planning results were displayed. For instance, below is an example of a Planning Book mockup in SAP SNP, which does in indeed show weekly buckets.

However, this is not in fact how orders are placed. Orders are in fact placed on days, which are scheduled from the requirement dates. In most planning systems, display is distinct from the order logic. In fact, in SAP SNP or DP multiple periods can be programmed to show in the Planning Book view. That is the interface can begin by showing the daily, then the weekly, and then the monthly values. This is called a telescoping time bucket, and how to configure this is described in this article. None of this effects the actual order dates. Supply planning in weekly buckets is unnecessary and probably counterproductive because eventually the purchase orders and production orders must be placed, and they will be placed on days. This is discussed in detail in this article.

On the topic of general planning detail, I decided to follow up with two people I know at ToolsGroup, Jeff Bodenstab, the VP of Marketing and Girish Bakshi, their Director of Consulting. They both echoed Joseph’s points from our earlier discussion, but added some extra information to Joseph’s comments.

Two Examples of How ToolsGroup Plans in a Detailed Way

They stated that this orientation extends in several areas of how ToolsGroup models the supply chain, two good examples being the way they look at demand history and how ordering batching is performed in the ToolsGroup application (SO99). They stated that when companies look at the daily line level, patterns begin to emerge that are lost when this same data is aggregated. In their analysis of the data of different companies, they have observed a tendency to the data points to cluster around day, week and the month. For instance, order volume is generally not evenly distributed within the five workdays of the week. Capturing this information and not losing it to averages is very important. Furthermore, this relates strongly to the topic of demand sensing, which is currently a very popular area of discussion in supply chain planning. The essential concept of demand sensing is that instead of looking at demand in broad-brush strokes, a more detailed analysis is taken.

Another observation brought to light by ToolsGroup is how they see the growing acceptance of this concept of detailed modeling. While SO99 had been designed this way from the beginning, in the past it was primarily ToolsGroup bringing this approach based on its internal philosophy to companies. However, more recently companies have begun to prompt them as to whether their software could model in this way. ToolsGroup is seeing the market increasingly come along to this way of thinking. I found that quite interesting because I work in environments where there is really too much aggregation being performed and accepted as a good practice. For instance, I would say it’s more common that not for forecast error to be declared and measured in terms of an aggregate above the product location. The most inappropriate forecast error level I have ever been quoted was an error based upon monthly sales dollars, and this was not an error generated for the purposes of the S&OP process or for budgeting.

Supply planning requires a forecast error at a product location combination. This is because the product location combination error drives, or partially drives, the safety stock calculation. (Other costs to forecast error include redeployment costs, stocking out on products that could have been stocked with the mis-forecast inventory dollars, spoilage/obsolescence, warehouse costs and several others. However the immediate implication for supply planning is increased safety stock. Forecast error measurements higher than the product location are not particularly relevant for supply planning, and as supply planning the primary consumer of the forecast. (For this reason I contend that the error for a demand planning department should be stated at the product location, but I will limit this elaboration here to refocus on the initial topic of this article.)

Conclusion

However while I still see far too much aggregation in companies, I only see the clients I work with, and it is interesting to broaden one’s perspective by taking into account what other people are seeing. And from what I learned from ToolsGroup, the appreciation for higher degrees of detail in supply chain planning is a very encouraging development.

Questions? 

Was this article clear? Do you have any experiences regarding the costs and benefits of planning at a detailed level? If so comment below.

Share

Tagged as: ToolsGroup