With shipments arriving daily at a factory’s dock, who has time to argue with the supplier, even if their materials do not meet the manufacturer’s specification? The only way to determine if the inbound materials meet the specification for each lot is by either testing the incoming materials or by reacting when a problem shows up at the end of the manufacturing line (or even worse, from a customer complaint). Certificates of analysis (COAs) hold the key to improving quality and productivity and focusing on continuous improvement rather than just covering your assets.
Explore this issueFebruary/March 2005
Issues with Paper COAs
COAs provide a snapshot into a supplier’s production process. The piece of paper that arrives with every shipment tells the manufacturer the lot characteristics of that batch. With the test results on a piece of paper, there is little that can be done with a COA for the following reasons: Static Information – Information on a piece of paper cannot be used to do anything else other than provide insight into a single point in time. Paper COAs become just another document requirement when receiving the materials. A single snapshot does not tell a complete story of a supplier’s ongoing compliance with a manufacturer’s specification and does not reflect possible variability. Difficult to Manage – Pieces of paper tend to get misplaced, lost or damaged. How many times have you asked your suppliers to resend their COA documents? How long does it take to retrieve the COA in an emergency situation (bioterrorism audit, line shut down, product recall, etc.)? Difficult to Share – Even if the COA is shipped ahead of time, the document is unavailable to multiple individuals unless there is a concerted effort or system in place to share that information. Purchasing, quality and production all have an ongoing interest in seeing how well a supplier’s materials meet their specification. With paper COAs, information sharing is very difficult.
Costs Can Be Steep
The following real accounts come from various process manufacturers who received materials from their suppliers, all with COAs, all within specification, and all resulting in huge costs. A consumer goods manufacturer recently had to take back a significant order from a leading retailer because the door colors did not match. The cause was the plastic pellets used to make the doors came from two different suppliers. Both incoming shipments met the specification, but were at opposite ends of the specification, resulting in different colored finished product. This had disastrous results from end users, retailers and manufacturers perspectives, costing well over $100,000 in returned goods. The only winners were the shipping companies. A cigarette manufacturer recently found that one of its suppliers was sending materials that met specification on paper, but upon further COA analysis, found that they actually were out of specification. This caused their machines to shut down at a cost of over $2,500 per shut down per line with over 20 lines, as well as having to run slower. This cost them real dollars in yield and productivity. A consumer packaged goods manufacturer regularly shuts down its high speed manufacturing lines because a container that they use frequently begins to drift out of specification, preventing effective accommodation of cans. All the containers that do arrive meet the specification, but tend to move from one end of the specification to the other. Tracking COAs can help spot trends so that they can effectively adjust their machines, as well as communicate with their supplier that they may have some problems in the near future.
Cause and Effect
The new COA has a many unique characteristics. Each measurement is in a format that can be used in a report. For instance, instead of receiving a paper COA, the COA arrives in a spreadsheet, electronically collected by a centralized system or sent via a message set along with the advanced shipment notification. The supplier sends the COA in advance of the material arriving at the manufacturer’s dock. Since the COA is in an electronic format, the data can be extracted and used by other programs. The COA data can be shared across the organization. It is possible to tie output effects to input raw material characteristics, and do this in real time. By having the data in an electronic form, in advance of it arriving at the dock and available for analysis, you can determine the cause and effect of different variables on your production process. By doing this analysis, you can then get alerts when those conditions are present (both good and bad), eliminating issues that reduce yield or quality. You can also set your production characteristics to get the highest yield possible. From a financial standpoint, the definition of a good batch and a bad batch is distilled to how well an individual batch contributes to profitability. If all batches are ranked from the best to the worst, (see Fig. 1), a small percentage at the top is great and a small percentage at the bottom is terrible. If we plot the P&L contribution for each batch, we get a strong correlation between contribution and the ranking of the batches. If we could understand the cause and effect between a good outcome and a bad outcome, we could identify conditions that produce a desirable (or undesirable) result. We could make a lot more great batches, stop making terrible batches and improve many of the batches in-between. The result is improved profitability (see Fig. 2).