10 Steps in SKU Profiling

Jan 31, 2007 8:32 PM  By

For a long time, finding the most efficient method of storing and retrieving inventory in the warehouse was quite a challenge: SKU data was very difficult to gather and even more difficult to process. Now, however, technology has developed to the point that anyone with the right inventory tracking system and a PC can conduct at least a rudimentary analysis of their products.

But when using the results of a product assessment, you must recognize and remember the potential shortcomings of your analysis. Following the 10 steps below will help ensure that you store product in the right type of storage and picking media while also having enough locations dedicated to each SKU.

1) Set a goal–and stick to it.
People profile their products for a variety of reasons. You may want to determine the cost of or justify the purchase of new equipment. You may want to fine-tune existing systems and equipment. Your goal may be to reduce picking and/or replenishment labor. Or perhaps you are simply trying to make the processing of the 20% of SKUs that make up 80% of your lines more efficient. Whatever the goal, it should always be top of mind when you perform an analysis. Getting off track to explore unrelated information can be a large waste of time.

2) Find the truth.
Getting data that truly represent what is going on is sometimes the most difficult step for companies. First, the data must be gathered from the right source. For example, data from enterprise resource planning (ERP) software will not necessarily be the same as data from warehouse management system (WMS) software. Just getting the data isn’t enough, either. Remove orders or lines that are not valid or were not processed using the system. The amount of data collected should be representative of your business climate and the frequency with which you are going to perform this analysis. Keep in mind the less obvious issues that affect spikes in your product demand, such as the amount of seasonality, product introductions, and promotional giveaways. Getting cleaned-up data is the only way to know the facts. Bad data will lead to bad results every time, regardless how thorough your analysis is.

3) Educate yourself.
Research the marketplace to select the right tools for your operation. There is a wide variety of types of storage and picking media, all having different characteristics. Do you want the operator to travel to each pick location, or would you rather have the product to be brought to the operator? Do you have consistently sized products or a broad range of sizes? What requirements do you have from a lines-picked perspective? Do you need a tool flexible enough that adding resources can increase picking output? How flexible does the equipment need to be regarding reprofiling items? Can picking and replenishment be performed concurrently?

4) Make it fit your system.
Now that you know the potential solutions available, you must determine which ones you can incorporate into your business. Not all solutions are adaptable to all situations. If you require strict lot control in picking items or follow FIFO rules, this could restrict the type of storage media you consider. Does your system bring the tote/shipping carton to the pick location, or is the order picker required to retrieve it? Do you pick all your items as full cases, all as split-case picks, or some combination thereof? Are these different types of picks going to be performed in the same physical space? If they are going to be separated, your analysis should also be separated by type of pick, since an item that is popular as a split-case SKU might not be a popular full-case SKU. If you are planning on processing these separately, do your operations provide you the room to consolidate the orders downstream of picking?

5) Look into the future.
Historical data may not accurately reflect the future for a SKU or a product family. Demand for a SKU may increase much more than the historical data show because of the release of a new product. Adversely, demand may drop off because of a scheduled product retirement. Business units or acquired product lines to be brought into your facility data will not be reflected in the current demand data. You may have to work with other departments within your organization, such as sales, marketing, and research and development to obtain this sort of forward-looking information.

6) Determine replenishment frequency.
The amount of labor dedicated to replenishment tasks will determine how much of each product must be kept in your pick location. Dedicating more resources will enable you to run leaner, allowing for more SKUs in a smaller area. For some SKUs, you need to store a minimum number of units in your picking locations. This helps ensure that the product is there when needed. For other items, an average amount of demand can be kept in this pick location. Doing so will keep enough product in the location to cover most days’ demand, but it might allow a small blip to exceed what is available in the location. Your tolerance level for an out-of-stock situation and determination of an acceptable time period for processing an order are the deciding factors here.

7) Perform a reality check.
Step back and scrutinize your initial results. Computers are very good at crunching numbers, but they do not “know” your product, facility, or business climate as you do. Computers do not know if you want items that are always ordered together close to one another to minimize the number of stops while picking the order or if you want to balance the workload by separating those items. Software may assign an item, such as a nonconveyable or aerosol product, to an inappropriate storage media.

8) Plan ahead.
Think through the reconfiguration process step by step. Are you going to “create” orders to pick the balance of the item from each location, or will you allow the item to deplete naturally through the fulfillment of incoming orders? The latter might not be efficient enough if the item has had a downturn in demand. Does your software provide the functionality you will require to efficiently move the product? Do not assume that you can flip a switch and your system will be reconfigured overnight. Depending on the type of equipment, reallocating locations to new items might take some time in addition to the actual movement of the product. Unless you have seasonal demand in which you can take advantage of a lull, this effort could have a noticeable effect on productivity.

9) Perform the reconfiguration.
All of your analysis will be for naught unless the reconfiguration is executed. Putting it off until a better time may guarantee that it never occurs. For rapidly growing businesses, there is no better time than the present to increase operational efficiency to support that growth. Cutting out inefficiencies lowers your bottom line.

10) Analyze the results.
Prior to profiling your product, select some relevant performance metrics to monitor for a while. Then, after reconfiguring your products, analyze these same metrics to measure your success. While the benefits might be noticeable without these measures, it is always helpful to have quantifiable proof.

Alan McDonald is a senior consultant at Mason, OH-based Forte http://forte-industries.com/index.php, a supply chain consulting and engineering firm.