No Such Thing as a Perfect Product

This week, we have a report in LDM on the SMPTE conference in Hollywood from Chris and another report in MDM on the VR event in New York. There’s lots of other news as well.

One of the stories of the last couple of weeks has been on the topic of the problems of the OLED in the Google Pixel 2 and just as we went to press, I spotted a report that lawyers are circling Google and thinking about a class action. It got me thinking about an issue that often took a lot of time when I was on the supply side of the industry.

The reality is that few products are perfect, especially when they are analogue, or depend on software, which is rarely free of bugs. I worked in the hardware industry in the sales and marketing of printers and monitors, mainly, in the 1980s and I always remember my sense of outrage around the middle of the 1980s when, for the first time, I was selling Microsoft software.

A customer made a claim to Microsoft about a significant bug in a software package. Microsoft said that it would be fixed in the next release, so the customer should pay for the upgrade. In the world of hardware, that would have been unacceptable, but although I argued the case for my customer, Microsoft was unyielding. I learned that software was expected to have bugs.

Later, I was running a company supplying really high quality CRT monitors – the best in the market. Now, CRT monitors were never perfect, as the magnetic components interacted in very complex ways and all affected the colour purity, white uniformity, geometry (distortion and linearity), convergence and focus. Fix one thing and it was likely that you made another problem worse, so the factory spent a lot of time (with human operators) to come to a good balance. They also were in constant feedback communication with the component makers to fix consistent issues.

One particular customer had bought a monitor and decided that it wasn’t good enough. We got the monitor back and looked at it and we thought it was pretty good and told the customer that it was within specification. He kept complaining, so I invited him to come to our warehouse where we had a factory-trained engineer who did sample testing at the goods inward stage. We said that when the engineer tested sample of that model, the customer could pick a different unit of the same model, if there was one that he liked.

The customer spent half a day sitting with the engineer who showed him what he was checking and after seeing a lot of units, he decided to return home with his original unit! (I had offered him his money back, earlier, but he said that “nobody makes anything as good as your monitors”).

In the end, what that customer wanted was a perfect product, and we simply couldn’t make them. Nobody could. There were too many variables and analogue components. However, we did win a loyal customer for the future by trying our best.

(I chose my monitors carefully in those days. I watched our head of sales, who used to change his demonstration stock regularly, with the best samples coming in through goods inward. When he held onto a unit for a while, I knew it was a good one and would claim it for my desk!). In those days, ‘golden samples’, hand tuned for the best performance, were commonly made for sending for product reviews in magazines. This frustrated us, as all of our monitors were carefully set up. Others could make one ‘golden’ unit that was nearly as good as our typical products. Unfortunately, magazines and reviewers didn’t see the average, they only saw the special one.

In my Display Monitor days, I was given a “golden” 16:9 24″ CRT monitor by Sony that had fantastic performance. Once I had reinforced my desk to support the 50+kg weight, I learned to love the monitor.

So, Google has something of the ‘its not perfect’ problem. It’s OLEDs are probably better than almost any other displays available, but somebody, somewhere will point out that they are not flawless.

Happy Days!