On a cross-country flight this week, I read most of Smarter Faster Better, by Charles Duhigg, currently on the NYT’s top 10 hardcover/nonfiction list of best-selling books. It’s fair to characterize the reviews it’s gotten as “mixed,” and the more critical ones find it anecdote-heavy and light on specific guidance; according to these folks, it doesn’t deliver on its promise of revealing “the secrets of being productive in life and in business.”

I take that view as fair but, to my tastes, irrelevant. It’s not the standard I’d hold this book to, or almost any book, for that matter. Duhigg or his publisher may have felt obligated to make a claim like that one in quotes, and Duhigg is certainly a pro when it comes to knowing his way around publishing (he won a Pulitzer for his work at The New York Times and his earlier book The Power of Habit also hit the best-seller list). But it’s frankly a silly claim, not far removed from transparent (but possibly more harmful) puffery along the lines of “10 investment secrets of the pro’s.” So I picked it up with no such expectation of being magically delivered “in life and in business” from all the wasteful and feckless habits people who skip the book are presumably doomed to keep playing out.

And on those most modest terms I’m finding it opens new perspectives in my mind on timeless issues such as:

• Where creativity comes from: stories of the legendary comic fecundity of the early days of Saturday Night Live, of turning the screenplay for the animated picture Frozen from a dud to a megahit, how West Side Story came out of left field to redefine Broadway theater.
• Why some people excel under stress and others fail to catastrophic effect: the pilot at the controls of Qantas 32, the Airbus 380 that suffered a catastrophic, explosive engine failure climbing out of Singapore airport and returned to land safely with 21 of its 22 systems impaired or not functioning at all—”the most severely damaged airplane of its size ever to have landed safely”—vs. the pilots of Air France 447, the utterly airworthy and completely sound Airbus 330 that the cockpit crew put into what they let become an irreversible stall from straight and level flight at its cruising altitude of 36,000 feet and let dive right down into the Atlantic taking all 228 people on board with it.
• And, for purposes of today, how organizational cultures can be toxic or enlightened: The infamous GM Fremont (California) assembly plant where workers drank heavily (and worse) on the job and even actively sabotaged cars as they moved down the line, recklessly misaligning parts, leaving bottles and bolts inside doors to rattle away, vs. that very same plant’s (and those very same workers’) transformation under the leadership of the Toyota Production System into the most productive plant in the US, with the fewest defects.

Along the way—and here’s where we circle back to Law Land—Duhigg tells of a study under the auspices of the Stanford Project on Emerging Companies which took 15 years and ultimately included over 200 companies. (I’m chagrined and exasperated that I’d never heard about this work before.) The study was born in 1994 when two Stanford Business School professors, James Baron and Michael Hannan, got tired of students asking them to prove their repeated assertion that no matter how superior a company’s product or service and no matter how loyal its customers, the enterprise would eventually fall apart if its employees didn’t trust each other.

And while Baron and Hannan believed this fervently, they had in fact little or no data to back it up. So they began studying local companies, startups very much included, in the Bay Area where they could conveniently interview the principals if possible. Eventually they built a model including five categories of management style (and yes, of course, it’s a model and not hard-core fact; they would tell you themselves that some companies have ingredients from a couple of categories and that others morph from one to an other). The five types are:

• Star: Hires came from elite universities, were very well remunerated and given “huge amounts of autonomy.” VC’s tended to be very enamored of “star” culture firms, believing that funding the A-team held the highest promise.
• Engineer: “This is your stereotypical Silicon Valley startup, with a bunch of anonymous programmers drinking Mountain Dew at their desks; they’re young and hungry and might be stars some day but right now they’re focused on solving technical problems.”
• Bureacratic: Thick with middle managers, extensive job descriptions, org charts, and handbooks.
• Autocratic: Close cousin to bureaucratic, but driven by a single person, usually the founder and/or CEO, one of whom reportedly described it thus: “You work. You do what I say. You get paid.”
• And finally, the Commitment model. Firms where people could, even if most actually did not, work their entire lives. You might think this stodgy-sounding model would be antithetical to the Silicon Valley ethos, but Baron and Hannan found many examples. “Commitment CEOs believe that getting the culture right is more important at first than designing the best product.”

If we overlay Baron & Hannan’s pentagonal model onto Law Land, I suspect we’d find something very much like this:

• The engineer model is effectively the null set.
• The bureaucratic model contains almost zero inhabitants and those in this sector tend to have very short life expectancies or to emigrate away.
• The autocratic model characterizes some young boutiques led by one or at most a very small handful of charismatic founders; their prospects of surviving the wildness of their youth and adolescence is open to question. The half-life of firms in this sector, in other words, is relatively brief. They emigrate away or dissolve when they encounter succession planning.
• The star model is what many firms either are or desperately aspire to be.
• And the commitment model finds firms in quite rarefied company; it’s antithetical to lawyers’ powerful autonomy-seeking instincts, and requires profound and unwavering commitment to a “firm first” faith and mentality.

Baron and Hannan followed their start-ups for a decade. So what happened?

In short, about half ceased to exist, but a few of the other half became enormously successful: Perhaps no earth-shattering surprise here. What they weren’t expecting was how powerfully the management style correlated with success—even after correcting for all sorts of other presumably germane variables such as company age, size, access to VC funding, senior executive turnover, and the macro-economy.

As you’d expect, the star model produced some of the biggest winners: Putting all the smartest people in the room could indeed produce outsize results. But stars also failed in record numbers, and were #5 out of all five styles in making it to an IPO. (The professors’ hypothesis, which seems valid, is that star firms can succumb to infighting because everyone wants to be the star.)

Another of the five models, amazingly, outperformed every other management style in almost every way that mattered. Firms following this style, among other things:

• were fastest to get to their IPO;
• had the highest profitability ratios;
• by and large were leaner, with fewer layers of management and fewer managers at each layer—committed, long-term staff generally excel at being self-starters and self-directed, and they’ve internalized what the firm’s about and where it should be going;
• wasted less time on internal rivalries and internecine competition, because the firm’s interest trumped personal agendas;
• and were more committed to and more closely in touch with their customers, which served both to provide superior service and to act as highly attuned antennae sensitive to impending shifts in the market.

OK, OK, I won’t prolong the suspense: These were the commitment firms.

Now, some characteristics of firms managed via the commitment style are predictable, or at least unsurprising, and one can readily see how they’d support outperformance. Among them are: A high sense of trust internally and between the firm and its customers; generous investments in training (implying confidence by the firm that people will make the investment pay off by sticking around); high and spontaneous (not mandated or managed) levels of teamwork; and last but perhaps most importantly, what students of the dynamics of teams and industrial psychology call “psychological safety”—confidence that one’s ideas and perspectives will not be dismissed out of hand and that one’s status is not in jeopardy if the firm chooses to go in a different direction than you suggested or advocated.

“Psychological safety” is worthy of extended discussion on its own, but suffice to say nothing—nothing—seems to make real, consistent difference to team performance other than psychological safety. This is a strong claim, but Google, among other firms (and Google is into deep data like no one’s business) have discovered it on their own.

Team members can be diverse or homogeneous, fast friends outside the office or relative strangers, all introverts or all extroverts or a mix, all based in the same department or random departments, (you name it), and there’s no detectable correlation with results. Psychological safety seems to drive it all.

Still, I can hear some of you muttering that this touchy-feely stuff is all quite swell, but when the rubber meets the road what you really need are some high-wattage rock stars to take prisoners and get stuff done.

Baron and Hassan would beg to differ, and from a counterintuitive vantage point. I suspect they agree in their heart of hearts that talent carries the day, but how you locate, cultivate, and retain that talent is really the challenge, so we give them the last word:

Good employees are always the hardest asset to find. When everyone wants to stick around, you’ve got a pretty strong advantage.