For Book Club in January we returned to more of a “business book” theme and read Adapt: Why Success Always Starts with Failure by Timothy Hartford. This book was a fast and fun read (good historical references and story-telling, though maybe fewer conclusions for a reader to take away, which was less trite, but likely more realistic).
Adapt was written in 2012, and while the historical references are still relevant, many of those stories seemed a bit aged given current events. Mr. Hartford walks through changes in military decision making, budgeting and autonomy that led to positive outcomes in region-specific campaigns in Iraq vs. the command and control approaches of the pre-Petraeus regime, as well as the culture of innovation and rapid experimentation that the most successful companies at the time: Google, Facebook, etc. were actively pursuing. Fast forward 10 years and there are no more 20% projects at Google, and many of the free-wheeling innovation projects have been descoped and defunded in an environment of layoffs, and strategic growth initiatives.
And maybe that is the point in some ways: adapt or die. A key nugget I took away was a story early in the book. Mr. Hartford uses The Toaster Project to illustrate the interdependence in the supply chain as well as the vast complexity of manufacturing even simple components in the modern world. Basically one man tried to build a toaster from scratch and realized the insane levels of complexity required to produce a heating element, the plastic casing, etc. The point of the story is that our modern world is stunningly complex, but we are so engulfed in this complexity that we take it for granted. We are blind to it. We overestimate the impact any one person or leader can have because we fail to see how complex the problems are that current leaders face.
Because the system we live in is far too complex for any one person to understand, even experts in a particular area aren’t as insightful as you might expect because of the interrelatedness of things with many areas about which they know nothing. He uses ample research across many expert groups (analysts, hedge fund managers, etc.) whose expertise did not beat general market averages or predict outcomes better than non-expert groups.
His conclusion is the reason companies don’t stay at the top is often because they were relying on factors beyond their control to achieve their success. When those external factors change good management cannot sustain growth. One has to keep innovating, keep looking at the signals of success, and be tolerant to failure, rather than assume the people or initial thesis will persist. He calls that inability for people to continue to innovate once they achieve success survivorship bias.
So how do you avoid survivorship bias? You have to immerse yourself not just in your success, but you have to see all the failures that led to the eventual success, and you can never stop seeding new innovation. Optimizing cost models works well when everything is growing, but when markets contract, you need to evolve. Evolution strikes a balance between discovering the new and exploiting the familiar. The evolutionary mix of small steps and occasional wild gambles is the best possible way to search for solutions. Evolution produces ongoing “works for now” solutions and then builds upon those ideas.
What kills innovation: overthinking your ability to project perfection. See earlier comments about complexity of our modern era. Those who iterate quickly on many ideas, analyze the data to understand the successes, weed out the failures, and keep moving win.
It was a great discussion at book club, and a good break from the current rhetoric to remember that there are domains where brilliant people can drive impact and innovate. Some of the best companies of our modern era came from previous downturns and market contractions. I am excited to see what new innovative companies will spring up with AI tooling and the available talent pool. I would love to see Mr. Hartford do a revision of this book and see if his thoughts around decision modeling and analysis change given modern AI. It would be a fascinating read.