Countering Cute Storytelling

This post is authored by Samir Patel, founder of Askeladden Capital. It is intended as a companion to Ten Commandments of American Retail.

Dry, theoretical, textbook-style instruction is generally agreed to be suboptimal for learning about most topics; hands-on work – case studies, lab experiments, or so on – is a major improvement.

But it isn’t perfect! We’ll set aside lab work for now and focus on the softer sciences. As it concerns history, whether that of business/finance/investing or that of nations and ideas, all that case studies give us is one possible outcome of many – while in some cases you can come up with enough sample size to draw definitive conclusions, oftentimes, “major” events (whether strategic decisions by businesses or major empires) have such unique circumstances and limited replicability that you can’t have a high degree of confidence in the positive takeaways from studying them (this is discussed in-depth in Phil Rosenzweig’s excellent The Halo Effect.)

So first-level thinkers get blown up: “oh, Buffett did XYZ, so I will too.” Well, no, that opportunity set doesn’t exist anymore – so the somewhat more sophisticated second-level thinkers think instead, “oh, Wal-Mart built a successful business because the company had ABC traits and followed QRS process, so let me look for businesses that model those traits and that process.” And that’s directionally better than trying to plaster discount stores all over small-town America circa 2017, but we’re still not quite there yet.

Take a seemingly easy/simple takeaway like “be honest and have integrity.” This seems to be a commonality among leaders who are widely admired, from Benjamin Franklin to Abe Lincoln right on through Charlie Munger, and it feels like the sort of thing that should be universally true.

But is that always and everywhere the case, or are we not thinking deeply enough? Indeed, one of the few redeeming qualities of The Moral Animal (emphasis on “few”) was an interesting discussion of the multiple potential “stable” equilibria in some classic game theory exercise – I forget the details, but basically the takeaway is that in a world where everyone plays “tit for tat” and recognizes the benefits of cooperation, then that is the right strategy… but if you get enough people who will default to being mean all the time, that becomes the only workable strategy.

Obviously, “be a jerk to everyone” isn’t my takeaway from that – but the analog in the modern business world could be the question of: what’s the most effective way to read a company? I want the answer to be “kindness and puppies,” because I’m the kind of guy who is going to follow that approach damn the empirics, but unfortunately for me, what I want is not always what’s true. For every business that succeeded because of its cute, cuddly culture (Zappos, The Container Store, no okay look I swear this post isn’t just an excuse to stuff Amazon affiliate links all over my blog), there’s a clear counterexample of a CEO that reached previously unimagined heights by berating the !@#$ out of employees (Musk, Bezos – and note that I have no opinion on TSLA as a financial success or otherwise; I’m referring purely to technical accomplishments, of which it seems they have had many.)

These are contradictory and to some extent irreconcilable lessons – is the answer that leadership style doesn’t matter? That seems unlikely. It seems more likely that there’s nuance: in certain industries, under certain conditions, with a certain compensation structure and a certain kind of employee… one or the other is correct. Maybe. But are we able, from this small sample size, to ascertain in exactly what situations brutality is superior to empathy? Maybe you’re smart enough to determine that with a high degree of confidence; I’m certainly not.

So what, then? Abandon vivid, engaging stories and instead rely purely on dry, dull, uninteresting research papers that perform rigorous statistical analysis? Well, no, there are plenty of flaws in that approach too, as Nate Silver himself points out in The Signal and the Noise. There’s the issue of backtest overfitting to deal with, confusing correlation with causality, etc etc.

Here are a few simple rules I propose for getting the most out of reading stories:

Think probabilistically and make allowances for data which is not included.

Certain entrepreneur stories – Phil Knight’s Shoe Dog and Richard Branson’s Losing My Virginity come to mind – could be reductionistically expressed as, “I spent a lot of my young adulthood screwing around doing what felt right, and ended up becoming a billionaire.” It’s a compelling narrative, but also a dangerous one – it’s easy for someone who is a billionaire to look back and say ah, don’t be a corporate square, follow your passion and go on a mystical journey and everything will work out…… but I’m almost certain that if you had access to aggregate statistics, the life outcomes (in pure financial terms rather than more holistic life satisfaction) for 20-somethings who, at least at first, pursue serious corporate jobs to build a resume, experience, and a nest egg before branching out, are probably way beyond those who meander along without much direction and hope things will just fall into place.

I know some really smart and talented “high-potential” people from high school who, the better part of a decade later, are now taking that latter route, and financial stability ain’t on the horizon, let alone billions. (Yes, there is an irony in me saying this, given that I’m, uh, not and never intending to hold down a serious corporate job – but then again, I did do that for at least long enough to figure out what was going on…) More broadly, many entrepreneur stories include some episode where everything was mortgaged to keep the business going… often including the houses or credit cards of family/friends/etc. That doesn’t mean it was the right decision – it just means it worked out.

By reminding yourself that the story you’re reading is merely one example that may or may not be representative (and in this case is not), you’re better able to contextualize the learning and think in terms of, “doing XYZ might increase my probability of success”, rather than the directly causal, “XYZ -> success”. Attempting to read over a longer period of time (ex. covering the long-term history of retail dominance in America rather than just one isolated story) also helps to build context and figure out what was common to success or failure under different circumstances and environments, providing a higher level of confidence in takeaways.

Assign “inversion” takeaways a higher level of confidence than “positive” takeaways.

Every story has multiple sides: if someone wins, someone else probably fails (a competitor, an enemy, someone whose business gets disrupted under the same circumstances). Saying that “oh, X Y and Z led to successful businesses over time” is likely a weaker takeaway than “well, clearly A B and C led to failure in almost all circumstances.” For example, in the history of retail, businesses that focused on protecting legacy profit streams in the face of competition, vs. going full-throttle after new/evolving markets, generally tended to lose both their profit stream and their right to continue in business at the end.How could things have turned out differently – were there disaster paths that were avoided at the last second?

As another example, in Benjamin Franklin’s autobiography, many seemingly promising/intellectual friends of Franklin succumbed to vices like drink and gambling – which Franklin largely avoided. Similar to the prescriptions for a miserable life in Munger’s famous speech, focusing on how losers lost rather than on how winners won might be more educational, if less heartwarming. This isn’t perfect, of course, but it may at least help frame things in a useful manner.

Go for third-level thinking, not platitudes.

We’ve already discussed first-level thinking (direct cloning) and second-level thinking (cloning of what led to the approach, rather than the approach itself). In the case of retailers, the latter could be, for example – relentlessly focus on the customer. Going only one step beyond often leads to unhelpful platitudes, though. Third-level thinking is going even deeper – what are the little nuances people miss?

Amazon, for example, showed an almost alarming willingness to burn money going after new categories (to further its mission of providing its customers with what they wanted), but it seemed to do a very good job of quickly retreating in categories like jewelry, or formats like auctions, when it became clear that there were relatively insurmountable barriers to entry. Wal-Mart, for its part, had plenty of innovations of its own but also methodically borrowed every good idea everyone else came up with. So maybe the real lesson to build a successful retail business is: relentlessly focus on the customer in areas where you’re actually cut out to hack it, but leave no stone unturned in figuring out what works and what doesn’t.

Be aware of the author’s biases.

It’s important to remember that no book is an entirely accurate account of what happened – rather, it’s the author’s interpretation of the facts that were available to him, filtered by his schema. Obviously, an autobiography is usually going to present things in a more favorable light than a third-party biographer would, and this is not just for ego reasons, but rather because when you’re working more off your memory than objective third-party source material, you’re not necessarily going to have an accurate picture (as discussed in Schacter’s excellent The Seven Sins of Memory.)

That said, third-party biographers aren’t perfect either – often, they have their own agenda to promote, whether intentionally or not – one thing to watch out for is “man with a hammer” syndrome where an author has a certain strong ideological view of the world (say, an economic or sociopolitical view – think about all the crazy super-political people you know who are blinded to reality) and/or believes strongly in some intellectual theory (environmental determinism, etc), and they’re going to naturally (not maliciously) tend to emphasize the factors that support their theory and crowd out any possible explanations, due to confirmation bias.

As an example, Marc Levinson (The Box, The Great A&P) spends what seems like an inordinate amount of time focusing on regulatory and labor relations issues – clearly these were important issues in both stories, but at the same time, might there be more (or less) to the story than meets the eye? Or other angles that were equally important but got short shrift? Questions to ponder.

As another example from the reader’s side, given that I grew up in Texas and hold a strong philosophical belief in the agency (free will) and self-efficacy, I have a natural tendency to focus on actors and their choices rather than circumstances/context – which is part of the reason I try to balance it out and pay more attention to the context rather than the protagonist when reading business narratives.

Read more stories (critically).

I realize this is somewhat counterintuitive advice given that I’ve spent quite a bit of time talking about all the potential pitfalls of buying well-told stories hook, line, and sinker, but the more you read while thinking about these issues, the more you get a feel for quality “business journalism” compared to what I’ll term “too-cute,” “just-so” storytelling that is fun to read but not very substantial – ex. Malcolm Gladwell, Jim Collins, etc. One general heuristic I use is: how hard does the author try to fit the world into their preferred box, vs. acknowledging (implicitly or explicitly) the other potential explanations for what they cite?

Stories are here to stay, and almost certainly for the better – but it’s important not to confuse a good story with factual reality. I love well-told narratives about great businesses, and have learn a lot from them – time will tell if I learn the right lessons…