We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next ten.
So said Bill Gates, in one of the few quotes attributed to him that we can properly source. That 1995 comment, sometimes called “Gates’ Law,” sounds a lot like the more general “Amara’s Law” that is probably twenty years older , but did not use specific spans of time.
Either statement raises an interesting question - if we overestimate short-term progress, but underestimate longer-term change, is there a time horizon for when we’re likely to be right? Is there a point where the curves of our aspirations, and the larger world’s behaviors, are likely to cross? If so, should our planning efforts be focused on that point, at the cusp of enough change to be transformational, but not so much that we can’t really grasp its effects?
A four-year look-ahead has lately seemed to me as if it might be that sweet spot. That same amount of time keeps coming up in so many different contexts that I’m starting to feel as if it’s trying to get my attention, waving at me from different directions in the hope that I’ll eventually notice.
Examples of four-year planning cycles include the amount of time that’s notionally ‘normal’ for completing a college degree, the duration of a US presidential term, and what used to be a consistent and predictable span of time between successive Summer or Winter Olympics competitions. In every case, an individual - student or leader or athlete - has to get up on some morning, make note of a specific future date, then embark on a use of energies and resources that accomplishes certain tasks and creates certain capabilities by the time that calendar page gets turned.
Also common to these examples is the need to plan and maintain an optimal pace. Push too hard, too soon, and there will be a mismatch between ambition and ability. Take too long to start building foundational strengths and the moment when those strengths can best be used will suddenly be at hand before you’re ready.
Demonstrations through the decades
We can hand-wave the idea that there must be a crossing of the curves, but I’m specifically saying that four years might see the kind of real, but comprehensible, change that Bill Gates said we should not expect in either two years or ten. I’d offer two examples, decades apart but both demonstrating my point.
First, from 1972 to 1976, a change took place that I can put in both concrete and personal terms. At the end of 1972, the last exam in my high school chemistry class before the winter holiday break was not a chemistry test at all, but a test of our slide-rule skills. Our instructors did not want to limit our assignments and exams to the kind of problem that could easily be solved with no math aids, but it was not reasonable to expect that every student would own a scientific calculator – something that cost, at that time, about $400 (at least $2,400 in today’s money).
Start a four-year clock and...
Three years later, as a college freshman, I had to sign up for an hour of time on a programmable calculator in a shared lab space to do a small set of problems requiring iterative methods for approximate solution. We all had calculators by then, but a programmable calculator was still considered a rare enough thing that this topic was limited to only one or two lectures out of a semester-long differential equations course.
In the following year, however, a $200 calculator was not only programmable, but had “continuous memory” – meaning that it did not have to be completely re-programmed (either by keystrokes or with a magnetic card) whenever it was turned off and then back on. The course was now called “Differential Equations and Numerical Methods,” and techniques formerly treated as a special sub-topic were now routinely assumed to be (literally) at hand - from slide rule to handheld programmability in four years.
This is among the most impactful examples I can offer, in that it did not merely mean that we did the old things more quickly, but that we could now do things differently. People in a freshman class only one year behind mine may have notably different mental maps of how to solve an entire class of problems. (What’s happening right now with quantum computing seems a striking parallel.)
This is not as simple, though, as saying that things get better over time. Slide-rule reasoning may have built useful abilities for estimation and qualitative thinking. Lessening the need to build those strengths may have had some unfortunate side effects. In a 1994 paper entitled “Computer-Assisted Decision Making: Performance, Beliefs, and the Illusion of Control,” researchers found that people equipped with more advanced tools for decision-making had greater confidence in their results, even when their measurable decision outcomes were worse than those achieved by other test subjects who did not use such tools. Caveat calculator.
Is four years the look-ahead that we need right now? Perhaps it’s not a coincidence that we find this number in so many places. Perhaps that’s just the right period of time, enough to get things moving, but not trying to aim at goals too far off to be usefully pursued. It certainly suggests that an annual plan may ask for too much while actually achieving too little.
At the same time, though, when I’ve spoken during this past year to our Salesforce customers and partners about the need to accelerate, their strategies and efforts for digital transformation, rather than deferring or diverting them, I have yet to encounter pushback or skepticism. Rather, there has been remarkable consensus that this is the line of argument to take in dealing with many different stakeholders.
Let me therefore conclude with two questions: one obvious, one less so.
First: what can you imagine doing four years from now? We seem to be pretty capable of answering that.
Second: what would it take to make that all happen in 2021? The competitors you know, and the disruptors you have yet to identify, may be running their clocks at a speed that doesn’t give you those four years to get it done.
