By John Diebold,
Director of Supply Chain Consulting, Commonwealth Supply Chain Advisors
One of the most frequent comments we hear from our customers is that they are not sure we’ll be able to help them design their distribution center or warehouse layout because they have such “bad data” on their operation. Indeed, working with less than perfect data is a reality of the distribution center/warehouse layout process. Commonwealth has discovered over the years, however, that it is possible to develop a sound design and warehouse layout despite data gaps or inaccuracies by applying a few basic principles.
The context of the analysis involves analyzing product sizes and sales forecasts to determine the best method to store items in the distribution center and plan the warehouse layout.
1. Validate the data: Whatever data is available first needs to be carefully audited to ensure accuracy and identify what gaps exist. Typical steps in validation include:
- Confirm fields
- Check formatting
- Check consistency
- Check pack sizes
- Summarize data and review ranges
- Manage outliers
2. Incremental analysis: When creating warehouse layout design tools, follow a process of “incremental analysis”. Rather than writing complex queries, create more basic algorithms where the step-by-step results of a formula can be viewed and checked against intuition and known values. This methodology is vital to ensuring an accurate result. As each step of the calculation unfolds, counter-intuitive results can be identified and either the formulas or the data set can be corrected as needed.
3. Don’t let the perfect be the enemy of the good: At a certain point in the analysis, one must stop trying to correct or improve the data and decide that the materials at hand are “good enough” for analysis. Even if all of the required data is not in place, it may be possible to make some logical assumptions and continue the project. All assumptions should be clearly stated, and the design tool must be created in such a manner that in the future, the assumptions can be replaced with better data as it becomes available.
4. In the absence of data…survey! In some cases, there simply is not enough good quality data to create a reliable design. When this happens, it’s time to roll up the sleeves and survey the distribution center, collecting visual information on what needs to be stored. There are generally four methods which can be used to gather this information:
- Method #1: Product cube: Use a database of product dimensions.
- Method #2: Survey storage cube: Survey each item in each bin in the distribution center. Record the SKU number and capture some information as to the amount of space occupied by that SKU in the distribution center.
- Method #3: Utilize bin cube (bin-by-bin): Measure the dimensions of each bin, and capture the rough percentage of space in each bin which is occupied by product. This is not a SKU-specific method.
- Method #4: Utilize bin cube (aggregate): Measure the dimensions of each bin, and determine a general utilization factor that applies to each bin type. This is not a SKU-specific method.
5. Checks and balances: Even if “good” product cube data is available, it is almost always advisable to check the result of a Method 1 analysis by performing one of the other three analyses. The results should be relatively similar. If they are not, then an effort should be made to understand why the discrepancy exists. Is it due to a normal margin of error in one of the methods? Is an assumption incorrect? Reconciling the two methods will often provide valuable insights into the operation that would not be obtained otherwise.
Our whitepaper, “Confidently Committing to a Distribution Center Design When Demand is Unpredictable” discusses the topic of using “imperfect data” in the warehouse layout and distribution center design process. Click here to download your free copy.