Predictive analytics? No, Thanks.

Yogi Bera famously said (or famously didn’t say): “It is hard to predict, especially the future.”

In competitive intelligence, we devote 100% of our time to predicting the future. The other 50% of the time is dedicated to sending noise around. Yes, we give 150% to our organizations!

Beyond this, though, anticipating an unfolding future is critical for companies trying hard not to be the LAST to notice it. This is the toughest and the most exciting part of CI work.

There are many techniques to predict the future. Some are sophisticated and dumb (predictive analytics waiting for the Giant data), some are just dumb (big consultants’ reports or magical ranking which predicts the past). The problem is that many companies use these lagging indicators and call them leading indicators. True leading indicators are always small data of significance. ?

The term significance is significant because only the analyst and the user of the intelligence can determine what is significant. And it starts with the competition analyst using some proven methods to imagine possible futures. Examples include scenario development (taught by the venerable Prof. Helen Rothberg at the Academy), or hypothesis formation and testing using structural analysis (taught by the amazing Heather Hallenbeck at the Academy.)

And then there is my method. All rights reserved, patent pending.

The Backlash Principle

Early warning is based on the principle that change, unlike cosmic supernovas or earthquakes, starts with tiny data way ahead of the cascade but these signs tend to be ignored when you wait for Giant Data. When using Giant Data and Predictive Analytics, your company is already late to the party, sometimes by a whole decade (e.g., Microsoft's Edge, CNN +/-, Amazon Fire phone, etc., etc.) That ensures your organization will forever stay reactive.

My ultrasophisticated method of predicting the future – a derivative of quantum computing backed by a platform (my favorite is a Sharpie platform but sometimes I use a basic dry erase) - is to mentally count. Yes, count.

I “count” responses on social media, items in the news, comments on LinkedIn, tallying “likes” and “angry” icons in my MSN morning feed.

Accuracy is not important. Gut feeling for “things are changing” is.

One tiny clue to change is to watch out for backlash long before it is obvious. Backlash is an important concept in physics, politics, and competitive intelligence. The unavoidable Newton’s third law in physics: When one body exerts too much force on another body, the second body will at some point push back. In system theory, it means pushing too far always has unintended consequences. It is already obvious in the current anti-Woke and parents-have-some-rights rebellion taking place.

In competition analysis, in turn, it stands for don’t poke the bear.

Or, if you poke the bear, assess, estimate, and predict the backlash.

Example from a recent war game

I am running a war game with an Asian company that is contemplating a move into a new crowded segment. One can point out the consequences of a backlash in various formats (simulation, P&L sensitivity analysis, BOTE, etc.) but I focus instead on predicting who will push back, what kind and how hard the expected backlash will be. For that I coach teams to use a Bayesian prediction - given what we know about this competitor- based on Porter’s famous 4-corner model- here is what to expect. A Bayesian-type thinking allows a company to choose to poke or not to poke, how hard to poke, where to poke, etc. Only with this type of “most likely” analysis based on behavioral economic profiling of competitors can a company make a well-thought-of move.

Note that no method prevents surprises. Surprises are surprising because they surprise (a deep thought for today, copyrighted.)

Some moves, however, should not surprise. That’s where a competition analyst is worth his or her weight in gold. Oh, OK, in California it will be a tank full.

Surprise is in the eyes of the beholder?

Early warning is about small data that competition analysts glue together before everything is obvious. Do not wait for Big Data. If you can’t skillfully anticipate a “most likely” future, it means you are not doing CI, just collecting/distributing information. Not every future is worthy of contemplating and the future is far from being utterly random. That’s why I favor Bayesian thinking.

Alternative perspective: Ready for “anything” means understanding nothing. ?

Imagine if you will a management meeting back in 2008/9 in which you stood up and said, here is my prediction based on economics 101 (or maybe just economics 01): The artificially low interest rates and enormous QE will lead to inflation (which is a form of backlash from consumers and producers where too much money chases too few goods.) Let’s start thinking of either a very premium product (typically immune to inflation) and/or a lower-priced, inflation-resistant product before it’s too late.

Management may have laughed you out of the room then, but in 2022 you’d be made SVP Strategy. ?

Join us June 6-10 in Boston to learn and practice Bayesian thinking. Here is a Bayesian prediction: You will most likely thank me for changing your career.