Doctor's Note (October): Supply Chain Knowledge: Don't Reinvent the Wheel

If you ask the management in any company, the chances are that they will say that one of their most important assets is the knowledge of their employees. Yet most companies have no idea if their organizational knowledge went up or down last year. In fact few, if any, companies actually measure their organizational knowledge.

Supply chain planning deals with uncertainty, and how to make consistently good decisions as conditions change. This is not something that can be codified completely in computer systems and procedures. (Computers don't handle uncertainty very well). However, good supply chains codify as much of the relevant experiential knowledge as possible so that each crisis can be approached in a systematic way.

The sad truth is that much of the accumulated knowledge is with people and is lost when companies reorganize or trim their staff as has happened in the last year. To avoid this happening again and again, some of the better run supply chains have built a data repository to collect and maintain supply chain planning information. They do this for two reasons:

1. Supply chain planning is inherently data intensive and people need tools to sort through the data to isolate the relevant information. Summaries and aggregation of ERP data that is not always easily and directly accessible from the corporate data warehouse. The supply chain repository acts as the buffer and helps to insulate the data warehouse from user intensive requirements. Judiciously examine the inventory targets so that you can respond quickly. Often this is best done by positioning purchased parts and raw materials.

2. Most databases in a company deal with storing data about things that have happened in the past. They do not capture and maintain the exception rules and the exception procedures for systematically dealing with issues, nor are they designed to keep planning scenarios and options. When generating a plan, a number of different scenarios are usually generated and compared before a final plan is decided upon. The discarded versions of the plans are usually in the corporate data warehouse, but can be useful for later inspection if circumstances change.

Purists will argue that the corporation should have a single data warehouse. However, this is impractical. The corporate data warehouse is normally the "system of record" and needs to operate with auditable rules that maintain data integrity. The planning data repository (sometimes also called a data mart) is designed to extract data from the corporate systems and present it to users easily. The only data that it maintains is the data that does not belong in the corporate data warehouse.

The proof is of course in the proverbial pudding. Our experience shows that the companies that use a repository to maintain some of the organizational supply chain knowledge weather downsizing and reorganization a lot better that the companies that don't.