I just finished eating dinner and realized that I won’t be going to sleep tonight until I purge this afternoon’s events from my psyche. The day started with a fascinating interchange that involved one of my primary suppliers. This would have been a whole lot more fascinating – and, infinitely less stressful – if it had happened to someone else and I was reading what they had written, rather than pulling an “all-nighter” myself. But, it wasn’t… And, here I sit scratching my head.

It started with a defective thermostat housing that, by itself, could be the foundation of its own column. But, it isn’t. This column is all about numbers. However, in order for those numbers to have meaning: in order for them to be relevant, there has to be some kind of meaningful context.

This story begins with a 1998 BMW 528i that was overheating. One of the reasons it was running hot was a “stuck” thermostat and a leaking thermostat housing. When the thermostat housing was removed we saw it was warped and recommended replacement. And, that’s where things start to get interesting. A few days after leaving the shop, the BMW returned with a continued complaint of overheating. We had done a fair amount of work on the cooling system and there really was no reason for it to run hot.

The coolant level was down, so we began looking for a leak. We found one at the bleeder screw where it seals to the housing: while not quite loose, the screw wasn’t sealing. The tech who worked on the vehicle was livid. He is meticulous in everything he does; Obsessive-compulsive to the point that everything is checked two and three times before a job goes out. That’s one of the reasons his comeback rate is almost non-existent and one of the many reasons he’s here.

He filled and bled the system, drove the vehicle and pressure-checked it until he was certain it was perfect and we sent the vehicle out again. A few days later it was back: still leaking, still overheating, still a problem. Inspection revealed evidence of the leak you can see in the digital image.

Techs can be fairly demanding. I know. I was one for the better part of my working life. My tech wanted a new housing and frankly, so did I. This particular vendor has what an IT guy would refer to as a “robust” Web site which is filled with lots of choices. When I went to the site and looked up the application I noticed something of particular interest. There were two parts listed for the same application: two parts with the same part number, the same MSRP, separated only by an $8 difference in cost. As it turns out, that wasn’t the only difference. One was plastic and the other metal.

We had evidently purchased, and then installed the less expensive housing on the 528i.

I didn’t order the part, so I don’t know why one housing was chosen over the other. I’m not even sure my service adviser noticed there were two parts with the same number or that they were priced differently. He’s on vacation so I can’t ask him. However, I do know what it feels like to have the same car in for the same problem three times in as many weeks, and I don’t like it. So, I started asking lots of questions.

When I asked my sales representative how many of the plastic housings had failed, he responded almost instantly: 85 out of 1,835. And, when I asked how many of the more expensive housings had failed he responded just as quickly: 83 out of more than 2,500.

I’m not a statistician. Nor, am I a research analyst. But, still, I was impressed! The numbers were right there… on the screen, at his fingertips. Numbers that could make a significant difference in my life, numbers that no one had ever mentioned were so readily available before this afternoon.

Would it have made a difference? You tell me. Our business is not all that different from yours. We accumulated what amounts to hundreds of dollars in lost time and product on this job, money and time that can never be recovered. More than that, we lost the confidence of our customer. Can you attach a dollar value to that?

There is only about a 1 1/2 percent difference between the failure rates of the two parts, but that 1 1/2 percent can mean a lot! It can mean not having to do a job over a second or third time, and depending upon what that job is: it can mean the difference between profit and loss for a day, a week, maybe even a month.

Confronted with a choice of product or price in the future from this vendor or any other, you can bet one of the first questions I ask will be: Do you have the failure rate of this part? Why? Because they are more than just numbers to me!