When the Experts got it WRONG!
In 2004, I was invited to speak at a meeting in Detroit, car country. Since none of my examples had anything to do with automobiles, I thought I'd "whip-up" a quick case study. I needed something about cars that also involved chemistry so I'd be able to pick up the meaning of what I saw quickly. So, I decided to look at catalytic converters.
Just as a reminder, the technology analysis process is a repeatable process for identifying and measuring trends in technology. The technology analysis process is repeatable and has several steps.
1. Define the area using the intake interview, Twelve Questions to Define a Project.
2. Build a Sufficiently Broad Patent (or other document) Database and Validate.
3. Categorize and Segment into groups of patents that help to isolate the information needed to tell the technology development story. the set of categories and segments is the "model".
4. Visualize and Measure. Once the model has been constructed, we can very rapidly build detail oriented visualizations for scientists, engineers, CTI practitioners and analysts and we can make the measurements that help distinguish one trend from another.
5. Summarize and Interpret Results to form a "story" by interpreting the visualizations.
6. Communicate the results in reports or presentations.
From the trends you can see product development options, competitor activity or inactivity, licensing options AND surprises that could indicate a Christensen-like disruption that is headed your way, but you need to run the process first.
Catalytic converters are part of the automotive exhaust system. A quick look on the Internet led to the system diagram from the EPA Office of Mobile Sources 400-F-92-007. The noted systems became categories for the analysis.
It turned out that the clean air regulations had stimulated patenting activity in catalytic converter patenting.
Since time was short, I decided to use IPC codes to define segments in this study and to compare patenting activity in IPC codes for the two years that are circled, 2003 and 1996. The question we wanted to ask was what had changed from 1996 to 2003.
For those not familiar with them, IPC codes are the "knowledge management" system the patent offices use to classify inventions. The codes I used with the 4-digit version like F06F.
To find codes where growth rates were changing, we made a measurement with a scale of -1 to +1 and built what I call a Vitality Plot. On a Vitality Plot, segments (IPC codes in this case) that were accelerating compared to the master database are on the right and in the red (+1). Segments that were decelerating are on the left and in the blue (-1). Segments that were not too different from the "norm" were in the center, yellow (0). On a Vitality Plot, the larger segments are near the top and the smaller are near the bottom. The scale is non-linear. Here is what the Vitality Plot looks like.
It will help if we add the IPC code labels to the chart.
When you think about it AND when you remember that it was back in 2003-2004 when this was done, the thing you see are about what you'd expect. Then, notice the small B60L code in the red region (green highlight). When we look into the code and view the titles, we see that the code is talking about hybrid vehicles. Remember 2003-2004? Hybrid vehicle were NOT the norm. They were barely on anybody's mind. Yet the signal for a hybrid vehicle is clear and unmistakable. It is not just one patent.
At this point in the meeting I postulated that hybrid vehicles were a coming attraction in the auto industry. After a very little discussion, one person from one of the Big 3 (in 2003-2004) stood up and said —and this is nearly a quote—
"You are totally wrong.
We tested those electric cars.
They don't work.
We have them all back."
Then, the person sat down. Of course, the meeting ended too!
Back at home, I was analyzing the presentation "failure" and I went back to the data to see what companies were present in the 1996 and 2003 segments. Here's what I found.
Based on what you see, can you identify the company that the person represented?
It is now 2010. Would it help if I told you that you are now a shareholder?
The heart of this story is that with MEASUREMENTS of technology trends, you can see well into the future — 3-4 years in this story. Isn't a warning of 3-4 years enough time for you or your company to see a disruptive signal and make plans for the change that is coming?
Analysis of technology trends can do the same for you in your industry.