page contents

I have been involved with projects to help reduce the proliferation of SKUs clogging up the database at client companies. It is much like the need to rid your home of clutter where you sell the excess through eBay or donate to Goodwill Industries International, Purple Heart Foundation, or the Salvation Army. In my case I am fortunate that we have moved frequently, and each time resulted in clearing out unwanted stuff.


There are simple and more complex situations that we find client companies confronting with SKU rationalization. We’ll begin with less challenging situations. One approach is to use the Pareto Principle or 80/20 rule. Sort SKUs assigning the high volume items to the top 20 percent A items, the next 30% to the “B” category, and the balance into the bottom “C” group. By the way, you are allowed more than A, B, and C groupings, and you can select break points other than 20/30/50. Each company is different. You may be surprised if you have ranked on the basis of quantity to find fasteners your “A” items, and jet engines your C’s. A preferred ranking may be on the basis of standard cost where jet engines rank as A, and fasteners as B or C items. Also important is to review the ranking by average selling price and margin. There is value in ranking items by key customer, channel, region, or sales person.


Next step is to perform an XYZ ranking. Here you are looking at frequency of sales orders over the past 52 weeks or 12 months to determine which the most popular items are by demand. Items that are purchased very frequently are ranked as X items, those infrequently are listed at the other end of the scale as Z items. Here too each client establishes their own parameters. It might be that sales in 38 or more weeks are X, 5 or less weeks are Z and in between are Y.


It is helpful to create a matrix with ABC on one axis and XYZ on the other, listing the SKUs where they fall within the grouping. A picture often helps determine potential candidates for removal. However, there is always an important consideration. If you have a SKU that is a contender for dropping but is helpful to close a sale on an A type item, then you keep it. Important too is to run a valuation report of items in inventory that are considerations for removal. A small investment could be an argument to keep the inventory. Do you scrap these products and take the write-off or hold them with an indicator in the system flagging not to reorder or re-manufacture? Then too these could be subject to removal during next year’s review.


I have experienced situations where SKUs exist in the database where they have been replaced with a new and improved product. Similarly if you are in a high fashion industry, with the introduction of the new seasons fresh and new merchandise, did you ever get rid of last year’s product mix? In these cases the company should run a supersession linking the history of the old to the new in order to provide a valid history pattern for forecasting the new item. Sometimes the old was not dropped from the system. Often you do need to hold onto the old items to support a sales analysis where standard cost and average selling price could be different from the replaced item. Generally after a three year period this historic information may no longer be pertinent or valuable and should be dropped from the database.


Here are two situations where client companies had significant proliferation of SKUs and what created this condition. In one case the company manufactured garage doors. Garage doors come in various types: swing out, swing up, roll up, or slide to the side. Doors can be made of wood finished with mahogany, redwood, or cedar, wood frame with foam insulation and plywood or hardboard skin, steel in different gauges, vinyl, and aluminum. There is also a choice of a one piece panel if not used for roll up, while roll up doors are made in sections hinged together. Sections vary between three and eight panels, depending on height. The final complication is that each door is custom made and varies in width and height in one eighth inch increments. The approach adopted by this client to satisfy the needs for their enterprise system was to treat each door as a Make-to-Stock (MTS) item. Each item in turn required a Bill of Material (BOM) and a Routing to control the process through the plant. After a new order was received it was near impossible and extremely time consuming to find an exact match in the database with all the available permutations. Consequently the planners created a new Stock Keeping Unit (SKU), BOM, and Routing. The database was literally infested with tens of thousands of never to be repeated SKUs.


The solution was to change the way orders were entered into the system. We adopted a Configure-to-Order approach whereby the customer at order entry time was lead through a selection of choices: style, type, material, height, width, etcetera, and the configurator assigned a sales order number to track the item through production to shipment. Naturally the BOM was linked to the required material with the appropriate dimensions, and the routing modified to suit the order. This opened up a simpler ability to provide an Available-to-Promise (ATP) date based on constraints within the factory processing current and future orders. In this situation a significant number of SKUs were wiped from the system. With the old style of processing, forecasting was impossible as all forecasts showed zero demand with only one sale of history to contribute to historical demand. In the prior situation there was no analytical codes to group like doors into families of product. Under the new business process, by stepping down a single level in the BOM forecasting became a meaningful exercise. Group codes were now in place. Planning BOM by product family became a reality.


In working with a bag manufacturer a similar challenge was presented. As the grocery store checkout you are asked “plastic or paper?” Bags come in all sizes and colors. Here too each order was a reason to create a new SKU, BOM and Routing. The change in form, fit, and function was seen as mandatory because with each run different artwork was applied to the bag. Many bags from a particular family of bags were identical, except for artwork. An improved solution was to separate the artwork from the generic bag, and let the plant apply the artwork based on the production order detail. The greatest benefit was a significant number of customers ordered identical bags, except for personalized artwork. This allowed for larger production runs and quick changes made to the printing process to cater for each customer’s current artwork needs. In this situation forecasting was more reliable and production productivity increased.


Often times it requires that we stand back, understand the challenge, and implement a fresh and simpler approach. Simple Solutions to Complex Problems®. A bloated database with useless information is not conducive to providing meaningful management decision making information.

Comments are closed.