[Cross posted at Future Tense]
“Never mistake a clear view for a short distance.”
Last month I had an opportunity to listen to Paul Saffo of the Institute for the Future speak at the CIO Magazine CIO|06 The Year Ahead conference in Phoenix. I was there as part of CIO’s Enterprise Value Award Process Review Board and as a facilitator for several of the breakout sessions. Paul was the MC for the 3-day event and his opening talk offered his rules for forecasting. They’re worth having handy if you find yourself in a position to have to make some bets on what might happen next.
Before sharing his rules, Saffo made the point that he thinks of himself as a forecaster not a futurist. In his categories, a futurist is an advocate for a particular future, while a forecaster is an observer trying to understand and bound the uncertainties generated by events and trying to frame the choices that might influence the outcomes. Saffo used the following image (actually his image was much nicer – this is from my notes, but you get the idea).
Rule 1. Know when not to make a forecast. Saffo made pointed reference here to Apple’s famous Knowledge Navigator concept video in contrast with Doug Engelbart’s Demo Video from 1967. I think what Saffo was driving at was the distinction between setting out a vision that will drive inventors and innovators on the one hand and recognizing that a salient event has occurred that opens up uncertainties that you ought to factor in to your planning.
Rule 2. Overnight successes come out of twenty years of failure. If you’re not paying attention, you’re going to be surprised a lot. This is where Saffo began to offer his take on the role of S-curve k inds of phenomena and how to account for them in your planning processes. Two points that I took away here. One is that there early stages of these curves is when you typically have the most leverage, if you can find a curve that will make it to the knee. Nothing terribly new there. The second, which I hadn’t thought about as much, was the difference in planning errors depending on where you were in the curve. I’m used to thinking only in terms of the tendency to overestimate how fast things will happen in the early stages of development. I’ve been less tuned in to the equally likely tendency to underestimate speed and demand changes past the tipping point. BTW, one of Saffo’s specific observations relative to this rule was that he’s paying more attention to Robotics as potentially the next big thing.
Rule 3. Look back twice as far as forward. Another quick bit of capsule advice about how to think smarter when you are dealing with exponential/logistics curve phenomena. This is a rule of thumb that captures the essential error in our tendency to think in linear terms about power laws. The change you’ve lived through in the last 10 years is a predictor of what you are likely to experience in the next 5. Douglas Adams captured this most memorably in his 1999 essay “How to stop worrying and love the internet.”
Alan Kay has talked about this in the context of why we’ve had more success at dealing with smallpox than with AIDS. If you are dealing with something that is operating on exponential terms, then the rate of growth matters as much or more than the slope at any instant in time. Given our tendency to project on a linear basis our tendency to over or under predict actually depends greatly on when/where you make that projection. With smallpox, the growth rate/infection rate is so fast that by the time you make any projection you are likely to be over predicting. With a slow growing epidemic such as AIDS, early stage linear projections will under predict. The corollary, of course, is that the surprise factor in slow-growing exponential phenomena is much higher.
Rule 4. Hunt for prodromes. Learned a new word. For you non-medical types, a prodrome or prodroma is an early symptom or leading indicator. This is William Gibson’s observation that the “future is already here, it’s just unevenly distributed.”
Rule 5. Be indifferent. Don’t confuse your desire for a particular outcome with its likelihood.
Rule 6. Tell a story or, better, draw a map. Trying to package your insights into a story (or scenario if you need to justify your consulting rates) helps reveal gaps, risks, and opportunities present in the events you are trying to understand. It can also help you get a better grasp on the potential wild cards. Saffo was more keen on the value of trying to find a way to capture your insights into something more graphical/visual. The value there is that those representations can help you highlight important relationships more easily and they raise the possibility of revealing ‘whitespace’ where you might find important opportunities to exploit or risks to minimize.
Rule 7. Prove yourself wrong. The essential wisdom of the scientific method. Understand and resist the natural human tendencies to believe. Be careful not to rely on a single element of strong information. Look for lots of pieces of weak information that collectively reinforce your insights. Your search for strong information should be for that one piece of evidence that proves you wrong. Look for the one thing that will make you look stupid if someone else brings it up after you’ve gone public.
It was a well spent morning listening to Paul, as was the opportunity to interact during the breaks.
Technorati Tags : paul+saffo, forecasting, s-curve, predictions2006