That day, when the DPMO report came out, the supplier's face turned pale.
I still remember two years ago, our production line was in chaos with delivery dates due to defects in a critical material. That time, we held a QBR (Quarterly Business Review) with that supplier, and I still remember how heavy the atmosphere was in the meeting room. At the start of the presentation, we showed this quarter's DPMO (Defects Per Million Opportunities) report, and the number was 6210. Do you know what that means? The supplier's sales representative immediately turned pale, and even our PM, who usually loves to joke, was speechless. That meeting, we discussed for a full three hours before barely finding a few directions for improvement. To be honest, without that "shock therapy," I doubt they would have realized how serious the problem was.
Where's the Problem?
A lot of times, we conduct QBRs like we're "turning in homework." Suppliers come to report a bunch of numbers, you listen and nod, and then next time you have another meeting, the numbers might be better, or they might be worse. But do you truly know the "why" behind these numbers?
To be frank, the core of a QBR is not "reviewing reports," but "solving problems." It should be an opportunity for you and the supplier to sit down, lay your cards on the table, and pinpoint production line pain points. If you just display data without in-depth analysis, how is that different from looking at a dashboard by yourself? What you need is not a pretty chart, but a clear course of action.
So the key is, you must make the QBR a "dialogue," not a "one-way presentation."
How to Do It in Practice?
To make QBRs effective, you need to do your homework and know how to "ask questions." Here are a few points I think are extremely important:
- Data must be "meaningful": The data you provide cannot just be a Cpk of 1.08 or a DPMO of 6210; it must be linked to "actual production pain points." For example, you could say: "This month, the wafers you supplied caused 15 tool downs in the Etch step of our process, leading to a loss of 200 wafers." This way, the supplier will feel the impact.
- Prepare a "question list" in advance: Don't wait until the meeting to start thinking. I usually, a week in advance, compare the abnormal incidents I have on hand with the supplier's quality reports from the past three months, item by item. Identify items where the numbers are "abnormal" or "continuously deteriorating," and then write down specific questions.
- Focus on "a few key issues": QBR time is limited; don't try to cover everything. I usually select 2-3 problems that have the "greatest impact" or are the "most urgent" for discussion. For example, if your wafer yield drops from 99.8% to 99.5%, it might seem like a small difference, but this could lead to you shipping thousands fewer products in a month; that's critical.
- Demand a "concrete action plan": The supplier promised improvement? Great, then "how will it be done?" "Who is responsible?" "When will it be completed?" "What are the expected goals?" All of these must be written down. If they just say, "We will optimize the process," that's just nonsense.
In other words, you must guide the supplier to dig from the "phenomenon" to the "root cause," and then formulate "quantifiable" solutions.
The Most Common Pitfall
The most common mistake I've seen is "fear of conflict." Many engineers find it awkward to have a falling out with suppliers, so they let issues be glossed over. Let me tell you, I've seen it happen too many times in my 15 years in this industry. Once, a new engineer at a QBR stood by and watched as a supplier pushed the blame for their own poor quality issues onto our process parameter settings. The engineer stammered and eventually let it go. What was the result? The problem remained unsolved the next month, and he got scolded terribly.
Frankly, QBRs are not for hearing good things, nor for making friends. They are for "uncovering problems" and then "solving them together." If you're not firm enough, not assertive enough, ultimately it's you and your production line that will suffer.
One Thing You Can Do Today
Before your next QBR, list 3 supplier problems you most want to solve and prepare specific data to substantiate them.