Using Predictive Analytics for Material Rationalization

Excessive inventory and lack of warehousing space is a common business problem. Rationalizing which materials stay or go requires analysis from different aspects of the business including sales, marketing and operations. SAP Predictive Analytics enables data from multiple sources and dimensions to be joined using correlation and scoring on a target field, we can recommend the disposition of materials. I’ve outlined the solution steps below.

Solution – Part 1:

The first step on the two step solution is to update the material data and apply mathematical calculations for forecasting, inventory reorder point and safety stock process.

1. Update material lead times based on actual receipt performance required for accurate purchase order placement.
2. Optimized the demand forecast using SAP Predictive Analytics to recommend a best fit model considering external correlated data with rankings of all forecast methods considered.
3. Update inventory strategy calculations for Safety Stock and re-order points relative to historical and future demand considering variability of demand over material lead time and proposed order fill rates.

Solution – Part 2:

The second step evaluates each material for obsolescence across multiple dimensions which is the core of the rationalization process. In the simplest form, sales dollars is the primary evaluation criteria.  However, using external data combined with sales derived data,  multiple dimensions can be evaluated and ranked on a material level.  This provides a more thorough picture of the material across multiple dimensions enabling all interested parties to agree on a material disposition.

The criteria used are:

• Recency – Compared to the start date of the analysis, the number of days since the last sale.
• Frequency – Over a defined time period, the number of sales order line items has this item appeared on.
• Monetary – Over the defined time period, the total sales for the material.
• Profit – Over the defined time period, the total profit for the material.
• Sales Indicator – Over a defined time period, percentage of sales increase or decrease.
• Number of Customers – Number of customers purchasing the material.
• Product Life Cycle – Develop a value for the age and expected remaining life of the material.

Ranking each material of the above criteria for provides a view for each dimension and the Material Total Score provides a relative ranking for all materials.   This analysis also supports using existing segmentation such as Material Groups.   SAP Analytics also can provide insight into segments not readily visible by using Cluster Analysis.   This technique identifies groupings or Clusters of materials based on selected or all attributes which can be useful in Marketing, Sales and Operations.

The Results

Using this analysis has historically provided a reduction of two to six percent of material Stock Keeping Units (SKUs)’s and provided a common discussion / analysis tool that Sales, Marketing and Operations can agree on and use into the future.

Summary

All companies require an analytical and methodical process that crosses multiple disciplines to optimize their SKUs and space. Presenting the facts, based on easy to understand mathematics, helps eliminate the emotion and focuses a difficult decision on the facts leading to optimized inventory budget and space. SAP Predictive Analytics is a key tool making this process easier and more efficient.