November 19th, 2007
Three trend analysis techniques
There are three types of trend analysis that I have used in the past to predict the future: geographic, temporal, and intuitive. I describe these three in the introduction to my Seven Trends in Networking and Security pitch. (coming to your neighborhood soon!) They caught the ear of a journalist in Canada recently so we did a short podcast on that topic instead of the usual the sky-is-falling stuff. He wrote the seven trends up as well.
Herewith the three types of trend analysis:
1. Geographic trend analysis is the easiest and most reliable. Simply visit a country, region, city, where something is taking off and predict that that trend will spread, if not to the entire world, at least to some other area. Back in the 80’s many an automotive pundit visited Tokyo and came back to Detroit preaching the gospel of small, fuel efficient vehicles. I launched an ISP, RustNet, in Michigan based on seeing the fast adoption of the Internet in California. And you may have read how my trip to Istanbul in 2004 exposed me for the first time to hackers that reacted to defensive measures taken by banks trying to protect user accounts from their keystroke loggers. Those are all examples of geographic trend analysis and prediction.
2. Temporal trend analysis can be misleading if it is applied to random activity or leaves out critical factors. Stepping up to a roulette table and plunking $200 on red because it has come up seven times in a row for instance. Everyone knows the chance of red coming up again is Less in that case.* Temporal analysis can be used to predict the outcome of US involvement in wars of occupation (long costly affairs with indeterminate results). You can easily predict the effect of “change” in Washington: more taxes, more spending, more regulation. Or the effect of new tariffs and import restrictions: higher prices, less trade, and a tumbling stock market. Thus, the past is prologue.
3. Intuitive trend analysis relies most heavily on the ability of the futurist to absorb all there is to know about a topic and then predict what the logical outcome will be. This type of trending obviously is fraught with the most difficulty. Many times there are external factors that really influence the outcome. Gartner is famous for predicting that OS2 would be the desktop of choice for the enterprise. The analysts involved in that prediction were too close to the technology. They lost site of the market dynamics: super cheap software on even cheaper hardware. OS2 was indeed the better operating system for the office environment.
I am putting together my predictions for threats in 2008 using these techniques. But first I have to take an honest look at how I did for 2007. But, that should wait until December. There is still time for some of these to come true!
*I don’t really believe that red has less of a chance of coming up just because the ball has landed on it seven times in a row already. The chances of red hitting are 18/38. Period. (Assuming two greens).
Richard Stiennon is an industry consultant. See his full profile and disclosure of his industry affiliations.
