About four years ago, I ran my first 100-mile trail race.
I spent four years working myself into shape. In 2008, I weighed 210 lbs. On race day in 2012, I weighed 150.
Leading up to the race, I ran an hour a day, every day during the week, and then three to five hours every Saturday and Sunday.
I prepped everything I could on the mental side too. I read every blog and race report I could find about the race. I watched every YouTube video anyone had made about the race.
And while I was preparing physically and mentally, I planned everything I could.
I had an excel spreadsheet on my computer that estimated what my splits would be at each aid station. I planned scenarios for good days, mediocre days, and bad days. I planned what I would eat at each aid station down to the last calorie, and exactly how much water and electrolytes I would take in, too.
And then race day came.
About 30 seconds after the race started, I realized I was wearing the wrong shoes for the course. I fell for the first time twenty minutes into the race. And then I fell another dozen times in the first half marathon. It took everything in my power not to fall off the side of the trail.
By the time the first aid station came around, I was already about 25 minutes off my worst-case scenario pace.
Then, just after the aid station, I went up a hill and got lost after a missed a flag.
There I was, about three hours into a 24-hour race, and I was already an hour behind schedule.
I had spent hundreds of hours contemplating a range of possible scenarios. And all of them were wrong.
You can read the full race report here. Bottom line: a bunch of stuff happened that I didn’t expect. I struggled but I finished. It worked out ok, though nothing worked out as planned.
After the race, I couldn’t stop thinking about how much time I had wasted planning for the race.
I had always been a planner. And I had always considered myself rational and economical in how I spent my time. But never before had I noticed such a stark disconnect in the time I had spent preparing for something and the return on the planning investment. It made me reconsider not just how I planned for ultra races (an obvious a waste of time), but it made me reconsider the notion of planning any activity.
My thinking now, with limited exceptions, is that most long-term planning is useless.
With most endeavors, I try to limit any instinct I have to plan more than a week in advance.
It’s not that I think that all predictions and plans are useless, it’s that I now have come to believe that most plans created outside an area of expertise or without a rigorous model are no better than baseline guesses. And since most of my predictions and plans aren’t based on a sophisticated model or coming from a place of expertise, I’m usually better off not planning at all.
Stated another way, for every endeavor, there is a moment when the number of variables in a system is large enough or not carefully understood well enough, where planning beyond that stage is a complete waste of time.
That’s a moment I call, “the Maximum Useful Planning Point,” or MUPP.
In Nate Silver’s book, The Signal and the Noise, he writes about how much weather forecasters’ predictions have improved over the past 20 years.
But as much as weather forecasts have improved, the value of a long-term weather forecast still fades after a few days. Forecasts a day or two in advance are now highly accurate, but weather forecasts for ten days ahead are still of no value at all. After nine days, weather forecasters’ predictions are no more accurate than the baseline assumption of the average weather in a specific location on any given day based on prior years’ information. And these are the best models in the world. According to Silver’s book:
After a little more than a week, Loft told me, chaos theory completely takes over, and dynamic memory of the atmosphere erases itself . . . Once the atmosphere has had enough time to circulate, the weather patterns bear so little resemblance to the their starting points that the models don’t do any good.
Weather is a dynamic system, in that the behavior of the system at one point in time influences its behavior in the future. And after a certain period of time in any dynamic system, chaos theory takes over and our models no longer serve to predict with any accuracy. That’s what I call the MUPP. For weather predictions right now, the MUPP is nine days for the best computer models.
The point at which one’s predictions cease to predict better than the base case scenario is the MUPP. This would apply equally to any system where you might ask the question, “is planning for a an event X days away useful?” The farther you try to plan into the future, the less accurate your model, and the more dynamic the system, the shorter the MUPP.
Let’s use another example. In chess, the more sophisticated you are, the more you can anticipate different scenarios. And it is a common supposition among novice chess players to think that the more sophisticated you become at chess, the more moves you begin to think ahead. This is true up to a point. But there is a declining utility the farther you think ahead, because if your opponent moves in an unexpected way, the thinking ahead immediately loses all value. With 32 pieces and dozens or even hundreds of possible moves every turn, depending on the stage of the game, most moves by your opponent will render your long-term planning useless. This is why chess masters don’t simply sketch out their entire game plans ahead of time, because there are so many possible moves, that the odds of executing such a plan are infinitesimal. Of course, in chess, knowing what your opponent is trying to do is important, but what is knowable or even probabilistic to know is usually a few turns on the horizon. The MUPP in chess is usually a few moves ahead.
In an endgame situation, a good chess player can absolutely think more than a dozen moves ahead, but that’s in part because it’s possible to force the action in a certain direction. In chess, the MUPP is very situational. In early to mid-game, the MUPP is only a few moves ahead. As variables decrease toward the end, it might be 15 or more moves ahead.
In games and in life, the MUPP is dynamic.
What I’m ultimately trying to combat here may seem like a bit of a straw man. But it’s something that’s important for me to remind myself. And it’s something I see often enough – organizations or individuals that seek to develop long-term plans or predictions, many of which simply have no hope of corresponding with reality.
The habit of long-term planning pervades every organization I’ve worked with, including most startups.
I’ve spent a disproportionate amount of my life constructing plans. And much frustration has happened when my carefully constructed plan diverged from the messiness and uncertainty of reality. Which is what happened about every time I made a detailed plan.
But part of me thought it was irresponsible not to have a plan.
And then a few months ago, I read this article by tech giant Marc Andreesen about his tips for personal productivity.
His biggest tip for personal productivity? No schedule.
Let’s start with a bang: don’t keep a schedule.
He’s crazy, you say!
I’m totally serious. If you pull it off — and in many structured jobs, you simply can’t — this simple tip alone can make a huge difference in productivity.
By not keeping a schedule, I mean: refuse to commit to meetings, appointments, or activities at any set time in any future day.
As a result, you can always work on whatever is most important or most interesting, at any time.
Want to spend all day writing a research report? Do it!
Want to spend all day coding? Do it!
Want to spend all day at the cafe down the street reading a book on personal productivity? Do it!
When someone emails or calls to say, “Let’s meet on Tuesday at 3”, the appropriate response is: “I’m not keeping a schedule for 2007, so I can’t commit to that, but give me a call on Tuesday at 2:45 and if I’m available, I’ll meet with you.”
Or, if it’s important, say, “You know what, let’s meet right now.”
Clearly this only works if you can get away with it. If you have a structured job, a structured job environment, or you’re a CEO, it will be hard to pull off.
But if you can do it, it’s really liberating, and will lead to far higher productivity than almost any other tactic you can try.
This is, of course, exceedingly difficult. But it is also profound.
If you try to anticipate what the best use of your time is today and tomorrow, you’re probably going to get that right. But much like the weather, if you try to predict what the best use of your time is going to be ten days from now, you’re probably going to be wrong. And so by committing to a schedule ahead of time, you’ll have committed to a non-ideal use of your time. If you try to schedule your life three months, six months, or years ahead of time, you’re likely to be not just wrong, but colossally wrong. And your commitment to doing something other than best use of your time will work against you.
And this to me, is the strategic value of non-planning. Because most people overplan their lives, those who have both the discipline to work hard and the restraint to not overplan can seize opportunities that arise without notice and take advantage of the unpredictable, giving them a competitive advantage over those who are rigorous or inflexible.
Frans Johansson’s book The Click Moment, Seizing Opportunity in an Unpredictable World, focuses on this very point.
This book is about two very simple but highly provocative ideas. The first one is this: success is random, far more random than we have come to believe. The second is that there are a number of specific actions that individuals and organizations can take to capture randomness and focus it in our favor. The reason people tend to consider these ideas provocative is because success, we are often told, is a result of strategy, planning, and careful analysis. Luck, on the other hand, is a force that lies outside of our control. This book rejects these conventional perspectives and proposes a useful and compelling alternative.”
Success almost never comes as a result of long-term planning, but through aggressive exploitation of nascent opportunities, ones that cannot be predicted ahead of time. Most businesses and professionals follow rigid plans and routines in hopes of that hard work leading to success. But, as counter-intuitive as it might sound, the rigid planning and routines often eliminates the possibility of pursuing the best opportunities as they arise.
Mark Zuckerberg was a psychology major who developed a “hot or not”-type App in college, and it happened to develop into a business. Had he maintained the resolve of sticking with psychology he would have never created Facebook. Bill Gates dropped out of Harvard because he knew that he couldn’t wait until he graduated and still develop the first interpreter for the Altair. Larry Page and Sergey Brin dropped out of grad school to start Google. Steve Jobs dropped out of Reed College to get a job at Atari (and later built Apple).
None of these businesses was planned.
So much of what it takes to become a visionary is the ability to abandon a plan, change your mind, and seize an opportunity.
Most startups know this well and have systems to accommodate for that fact. Startup methodology guru Steve Blank is famous for observing, “No business plan survives first contact with a customer.”
Many of the smartest engineering firms are also averse to long-term planning. Most notably, agile development methodology creates schedules according to tight feedback loops, mostly based on cycles shorter than a weather forecast.
Experiment. Test. Repeat. Explore. Make lots of small, calculated bets. Don’t bother with the long-term, rigid plans.
It’s the methodology that’s worked for many of the best-run organizations in the world. But it runs counter to the way we naturally think. We like to believe we can know what we’ll do a year or five years from now. But the moment we pretend we do, we put an arbitrary anchor on our future that limits how we live.
The ultimate question is: when is it more strategic to have a plan and when is it more strategic not to have a plan? Or, stated another way, how do we figure out the MUPP in any given endeavor and how do we use that information?
My conclusion is this: unless there is data or evidence that shows there is value in planning more than a week or so in advance, that there’s probably no value in doing so.
That feels chaotic. That feels uncertain. That feels unstable. But if you can discipline yourself to work and focus in an environment that is almost wholly absorbed by short-term priorities and immediate feedback cycles, you’ll find yourself able to exploit a range of opportunities most organizations will never even consider.