There is no doubt that demand segmentation can help you bring clarity to demand planning and the overall supply chain planning process and lead to far superior results in terms of various supply chain metrics. At the core of segmentation is the understanding that one size will not fit all. 

If your business doesn’t yet have a tool or process to do demand segmentation then Arkieva can certainly help you get started, you’ll also be better served by reading the following blogs on demand segmentation before reading further: 

  1. Demand Segmentation Starter Guide 
  2. Can Demand Segmentation Improve Your Statistical Forecast? 

While the segments should ideally be completely different from each other and perfectly homogenous from the inside (Examples – All the SKUs that have a seasonality or SKUs that are sold to a set of key customers or SKUs that have a very high order variability), depending on the level of maturity of your demand planning process and in general that of the overall supply chain planning process, there are always exceptions. While you can set rules and automate, to a large extent, to determine key supply chain decisions (such as stat forecast model, service levels, inventory levels, replenishment policy, etc.) for different segments, it’s the exceptions that always spoil the party.  

Depending on the nature of an exception and its business impact, the cost of dealing with it without a plan may be significantly higher than that associated with having one. For example, an exception or a set of exceptions could lead to a bad forecast for a strategically important set of customers and not having a plan could have a significant long-term business impact or a set of exceptions could cause a serious spike in inventory which again could have a big impact on the business on various ends. It is thus important to have a process within the demand segmentation process to identify and flag exceptions and then drive action items to improve forecast and replenishment. 

So, what should a process to identify and manage exceptions involve? Below you will find some of the key steps that one should keep in mind: 

  1. Definition of exceptions – One of the primary considerations should be the benefit received or projected versus the time spent in setting up and reviewing exceptions. A very broad approach might result in too many records being flagged as exceptions which could then make the process of going through each one of the time consuming, ineffective and not really cost-effective. On the other hand, too narrow an approach may leave risks unidentified and lead to a significantly negative impact on the business. Clearly, it is important to strike a midway.   
  2. Alerts and Notifications – It is important to notify the appropriate stakeholders immediately of the exceptions and potential issues in a timely manner. The stakeholders could range from daily planners to top-level executives who can remove roadblocks if any are responsible for the issues. 
  3. Periodical review of definitions – While it may be prudent to start off with those that could have a big impact on the business, it is important to review and as the process evolves, add definitions that could have a lesser and lesser impact. Above all as the business grows or as times change there is always the likelihood of newer conditions leading to exceptions. 

As you can imagine and are probably aware, the applicability of exception management is not just restricted to demand segmentation or the demand planning process but the whole of the supply chain planning process. At a strategic level, rules can be defined to identify problems at an aggregate level and drive a strategic decision. For example, a need for an additional warehouse to reduce lead-times or an additional plant to significantly ramp up capacity or a significant reduction in the demand plan for a given set of products which could, in turn, be a sign of additional issues. At an execution level, near real-time data could be captured for an insight of the current situation, for example, an order progression analysis can help flag sales lagging behind the forecast, or an on-hand inventory report of finished goods can help flag inventories exceeding targets. Because the alerts can be near real-time, the business can plan replenishment changes to avoid longer-term and more serious consequences. 

Read More: Demand Forecasting Analytical Methods

Arkieva can help you take a data-centric approach to setup definitions and manage exceptions and trigger real-time notifications at various steps in your supply chain planning process. I’d like to leave you with the following quote and thought: 

“A good teacher must know the rules; a good pupil, the exceptions”, oh well, how about exceptions to exception then? 

Enjoyed this post? Subscribe or follow Arkieva on Linkedin, Twitter, and Facebook for blog updates