Not All Data is Created Equal: How to Pinpoint the Right Defects to Fix Your Supply Chain

Alright, let's talk about defect data. When I say "defect data," I’m referring to the data that shows where things went off track—where a process wasn’t quite under control. The type of defect data you collect should be directly linked to your strategic objectives. Let’s break this down with a practical example.

Say your strategic objective is to improve customer service. A common metric for measuring customer service improvement could be on-time delivery. In this case, your “defect data” would be all those late deliveries. So, your task would be to gather data on recent late deliveries—maybe focusing on the ones that were most delayed. You don’t need to go overboard; around 30 data points will do. Make sure these data points are reasonably representative of the total population of late orders.

This approach isn’t limited to just on-time delivery. It can be applied to other supply chain objectives too. If you’re aiming to improve cash flow, your defect data might be the items with excess inventory. If reducing purchasing costs is your goal, you might look at Purchase Price Variance. The key takeaway is that these defects signal that something didn’t go according to plan—a process was out of control. And getting your key processes under control is the first step in any continuous improvement journey.

Let’s walk through another scenario. Imagine that maintaining product availability is the top supply chain objective for your organization. Projects impacting product availability would naturally get the highest priority. Sounds straightforward, right? Not quite. Here’s where it gets tricky: every department might interpret this objective differently, each believing their project should be the priority.

For example:

• Purchasing might insist on new MRP software to improve raw material planning and avoid production stoppages.
• Logistics could argue for a new warehouse to increase safety stock.
• Quality might push for a new Lab Information Management System to speed up product release.
• Manufacturing might propose investing in expensive new equipment to remove a bottleneck.

When considered separately, each project could have a strong business case and a clear ROI. They all might even be data-driven, but were they looking at the right data? I’d argue they weren’t.

Instead of jumping straight to solutions and then retrofitting them into strategic objectives, we should start with the problem—through root cause analysis (RCA).

Let’s apply this approach to our scenario. We’d first examine recent stockouts and dig deeper to understand their root causes. With this mindset, it’s highly likely that none of the proposed expensive projects would have made a real impact. Often, the actual cause is something much simpler and less predictable.

Here’s an example from the field: One organization was facing an increase in material shortages on the production floor. Naturally, everyone blamed the material planning process. But when they dug deeper, they found that materials had arrived but were stuck on the receiving dock. Why? It wasn’t because the material was defective; the supplier had recently changed their name, and the new name wasn’t on the receiving specification. As a result, the Quality Assurance (QA) department couldn’t release the materials. The person responsible for updating the specifications had a backlog and hadn’t been told to reprioritize until it was too late.

Would management have guessed that this was the primary issue impacting a strategic initiative? Probably not. And that’s why root cause analysis is such a crucial first step. By understanding the actual problem, we can avoid costly missteps and focus on what truly drives improvement.

So, the takeaway here is simple: Start with the problem, dig into the root causes, and let the data guide you to the most effective solutions. It’s not about finding the flashiest project; it’s about finding the right one.