A few weeks ago someone pinged me on Twitter with a very interesting comment. He said that whilst he enjoyed learning about business from biographies as much as the next person, “reading books about weightlifting doesn’t shortcut actually hitting the gym, doing the reps, feeling the movement and developing the instincts.”
This was in response to my quip that “the business equivalent of ‘a month in the laboratory can often save an hour in the library’ is ‘a decade of business experience can often save a few months of biography reading.’”
But this reader is also right, of course. There is much in business that can only be learnt by doing. I would never suggest ‘read business biography’ to someone who was just starting out. Getting actual experience is more important the earlier you are.
Business is not like weightlifting, however, and it is not like weightlifting in some very interesting ways. It is those ways that makes reading biography so fruitful.
I thought it would be useful to explore the ways in which business is not like weightlifting (or Judo, or most sports, for that matter). My response to the person went something like this (edited for clarity):
I think the two are different domains. Here’s an analogy. Let’s say that you move to a valley in a foreign land. You build a house with your bare hands, learning the skills of a mason and a carpenter and an electrician and a plumber. You build a beautiful house. All is good for nine years. But because you don’t read the history of the area or talk to the natives, in the 10th year a flood devastates the valley, destroys your house, and you die.
The thing about business is that you have capital cycles, and those are like 10 year floods.
In 2022 the US Federal Reserve hiked interest rates at the fastest pace since 1982. This increase was mirrored by many central bankers in major economies around the world. As a result, there was a sudden rise in borrowing costs across the globe, and a commensurate tightening of global liquidity.
The rivers of cash that had been running since 2008 — that had gushed especially freely during the COVID years — were choked off. Capital had a cost again.
And then a whole bunch of startups started dying.
Not all of them died outright, of course. Many of them got acquired, or were merged with other companies, with the mandatory “we started X to pursue Y mission, and are excited to continue our mission at Z” press releases, after which their products became dormant. Other startups pivoted. Some companies — who had raised far more than they were worth but were run by wiser founders, cut burn aggressively. They had realised that they were not likely to raise at their previous valuation, perhaps forever.
Yellow line is global liquidity (with units on the left expressed in trillions), and is juxtaposed against the S&P500 (units on the right). Global liquidity calculated as a sum of the US federal reserve system balance sheet minus the US Treasury general account and reverse repo market, added to European Central Bank balance sheet, People’s Bank of China balance sheet, Bank of Japan balance sheet, and includes M2 of US, EU, China, and Japan. Red box shows when global liquidity crashed. Full script available here.
Concurrently, a whole bunch of venture capitalists saw their VC careers end during this period. Some are still running funds today (never able to raise another) but most unserious VCs have washed out by 2025.
Mark Zuckerberg, the CEO of Facebook, announced a ‘year of efficiency’ in 2023 to placate investors.
And across the tech industry, the layoffs began.
It is in this way that business is not like weightlifting. In weightlifting, as with most sports, you do not experience regime shifts. In weightlifting you do not have 10 year floods that can kill you.
It’s important to notice how delayed many of these events are. The Fed hiked interest rates in 2022. I am writing this in 2025. Companies have died or done layoffs for every year from 2022 onwards. More are likely to come; companies that raised debt and investors that operate in private credit have locked in rates for a number of years. For these market actors, the consequences of their unwise actions (if any) have not yet come home to roost.
The pace of these changes are quick when put on the page, but slow enough in real life to be disconcertingly subtle. The previous operating environment that we were in — that I cut my teeth in — started in 2009 and ran for 15 years, a period of low interest rates that we call the ‘Zero Interest Rate Policy’ era (or ZIRP, for short). You don’t notice these things without reflection. Worse, most of us pick up near-invisible assumptions about the way we work from what we learn in our 20s. If your 20s coincided with ZIRP, like me, you’ve probably been in an adjustment period ever since.
(I should pause here and note that while I sound very wise in this essay, I did not see this coming. I was as taken aback by the changes that unfurled as anyone else. Do not be misled by the tone here, which is an artefact of my writing style.)
Actually, let’s run through my experience as a prototype. I want you to imagine that you graduated into the post GFC world, say in 2011, and learnt how to work in your first job in the 2010s. Perhaps a manager takes you under their wing and tells you “this is how we do things.” More likely you watched how people and teams around you execute, and internalise the operating cadence of your industry. As the decade progresses, you change jobs; you continue to absorb ‘this is the way we do things’ even as you rise in your career. And while each company is slightly different, there is a fundamental stance — a set of underlying assumptions that everyone shares that nobody questions. These are things like “you can become rich if you are an early employee in a startup” or “the solution to growth problems is hiring more people” or “we pay lip service to tracking operating costs but actually I have no idea what things cost; let’s pay for another SaaS because software is always cheaper than people!”
(But actually we pile up on software and people because … see point one: we don’t know how much things cost).
In the wake of the post-2022 regime change, you might have noticed some of these assumptions fraying. The thing that isn’t salient about cheap capital — or perhaps that you really only understand when you live through it — is that cheap capital gets into everything: it distorts buying behaviour throughout the economy. It’s not just startups that can raise infinite rounds of capital. Easy money means:
I started my career as a software engineer, and still have great affinity for the industry. As a result, I’ve paid special attention to the folks who have written about the change in operating assumptions over the last few years. I’ve noticed that many of them are older, wiser, and often have had a taste of the before-times.
Here’s one, an excellent piece by engineering leader Uma Chingunde:
When I think of what happened in tech as we went from ZIRP to post-ZIRP, I often think of the kids' game of musical chairs. Everyone walks around a set of chairs while the music plays—and when it stops, you scramble to grab one. If you fail, you’re out of the game. In early 2022, the music stopped abruptly for the tech industry with the end of ZIRP. Funding, hiring, and, by extension, career trajectories all took a hit. Although the rise of AI (most notably marked by ChatGPT’s launch in late 2022) has kept one part of the ecosystem humming, the rest of tech now faces a very different reality.
During ZIRP, hiring was frenetic, but the shift to post-ZIRP triggered a rapid slowdown, visible even before public layoffs began. There have now been multiple years of layoffs, but the number of open jobs has only slightly increased, as seen in this graph, leading to a large pool of available candidates relative to jobs. A larger pool of candidates means more conservative hiring behaviors. Hiring managers can wait to hire someone who has already done exactly what they need vs. taking a bet on someone who has the potential. This puts pressure on the flow of people through the system.
The tech job slowdown disproportionately impacts two groups: early-career professionals struggling to enter the market and seasoned leaders finding it difficult to re-enter at a comparable level in case of a layoff or a break. Lateral career shifts are more complex than before. If you take this further, it naturally follows that moving up career ladders is slower since fewer roles are being created at every level. Instead, team sizes shrink, leading to scope and opportunities shrinking. ZIRP was the grease that kept this system flowing freely; its sudden demise has left us with a creaky, jammed-up system that lacks lubrication.
Chingunde then walks through some of the less obvious implications. One of those implications is that it’s harder to hide now if you are incompetent. Or, to flip this on its head: over the past 15 years there were a great many senior people who were successful, who were paid high comp packages, who grew large, respectable followings and did shiny tech talks … who were actually subpar operators with good political instincts. This is a little surprising to me, and it might be to you, but it shouldn’t be — this fact is a direct consequence of “there is less accountability when money is plentiful.”
Keep this observation in mind; we’ll return to it in a second.
Here’s another piece, this one from James Boyer:
A senior leader recently handed me a “gift”: a plan to hire more engineers. I paused. “Why?” I asked. “What problem do we have that hiring solves?” They looked at me like I had passed Go and didn't want my $200. But adding more people doesn’t mean more progress. That's a ZIRP-era assumption. What matters now—and what should've always mattered—is leverage. Outcomes. Impact per engineer. Or as Charity Majors once wrote, “every achievement has a denominator”.
Let me tell you something interesting about this. Shortly after ZIRP ended, I read The Birth of Lean, a history of the creation of the Toyota Production System. It consists of a bunch of interviews with the folks who invented TPS over the course of the 50s and 60s. When I first read the book I enjoyed it but was insufficiently reflective. There is one interview in the book that haunts me today, now that I’m rereading it with this regime change in mind. It is from Chapter 5, with retired Toyota executive Masao Nemoto. It goes like this:
I was once responsible for overseeing Toyota’s 10 production engineering divisions. In the first year, the budget requests from the divisions totaled more than ¥150 billion. I pared the total down to ¥150 billion and took our budgeting proposal to the executive vice president (EVP) in charge. Here’s what ensued.
EVP: I can’t sign off on ¥150 billion. Cut it 20%.
Nemoto: If we’re going to cut our production engineering budget 20%, shall we also reduce the number of model changes 20%?
EVP: Don’t be ridiculous. We do what we need to do. And you do your part with 20% less spending.
Nemoto: Do you seriously think that’s possible? If that was possible, we would have been doing it a long time ago.
EVP: That’s what your brain is for. That’s why we put you in charge of production engineering.
Nemoto: So I’m supposed to come up with a way?
EVP: That’s right.
So I went to work on coming up with a way to complete our assigned mission while trimming our budget. That budget was the result of a lot of careful planning and calculation and negotiation. So cutting it 20% would be no easy matter. When I was in charge of purchasing, we worked with suppliers in value engineering activities to lower parts costs while maintaining quality and performance. The idea occurred to me that we could use value engineering to reduce our investment expenditures. I gathered the division general managers to explain what I had in mind. They were predictably opposed to the idea. Here is an exchange that ensued with one of them.
General manager: You accomplished your value engineering results with vehicle parts. But production engineering is a completely different story. With parts, you start with drawings and put together prototypes. Next, you evaluate the prototypes and prepare revised drawings based on your findings. Only then do you move on to mass production. In production engineering, we only get one shot. We prepare the drawings and then make the equipment based on those drawings. If we get something wrong, we don’t have the opportunity for trial and error like you do in developing parts.
Nemoto: I see what you mean. And I figured that’s what you would say. But I’m determined to get the budget down 20%. That’s our mandate from above, and I want you to work with me to see if we can’t do something. I’m thinking that we ought to be able to use value engineering in some way to do that, and I want you all to go back to your divisions and give this some thought. Talk about it with your people and see what you can come up with.
They came up with a way, and cut the budget by 20%.
The first time I read that interview, I thought to myself “oh, how cute.” But upon rereading the book, I realise that I should’ve paid more attention to this interaction. In my entire career, I have never, not once, had a conversation like that. “You’re expected to do 20% more, oh by the way, we’re slashing the budget by 20%. Figure it out.”
If someone had said that to me, or perhaps if someone had said that to a manager I reported to, I think they would have thrown a hissy fit. They would have thought it impossible. My entire generation in tech cannot imagine getting more done without also increasing headcount. Nemoto and his General Managers protested, but they got it done. Why? Because they’d seen it done before. Because Toyota did not do layoffs. And because money was tight.
When folks talk about TPS today, they throw around words like ‘tacit knowledge’ and ‘process power’. But they don’t often talk about the path it took to get there. That path involves conversations like the one Nemoto had with his EVP, again and again, over decades — a conversation that I doubt many companies have had in the past 15 years. And I can already hear you saying things like “oh, but software is not like manufacturing; manufacturing is this and this while software is that and that.” But have you ever said this in an environment when the cost of capital was high and climbing ever higher? Have you ever had a boss who said “that’s what your brain is for, that’s why we put you in charge of engineering?”
I know I haven’t.
The overall point I’m making here is that if you treat business like weightlifting, you will periodically learn bad techniques. Perhaps bad is the wrong phrase. Perhaps ‘techniques that are overfit to the current capital environment’ is more accurate.
Let me rephrase this in a way that hits harder: everything that emerged as a best practice in the previous 15 years is now suspect.
This feels a little extreme. If this makes you feel uncomfortable, fine — temper it appropriately. Perhaps you trust some recommendations and distrust others, and you have good reasons for your filters. But I think it’s worth taking this idea seriously.
Software engineering leader Will Larson has an essay that reframes this phenomenon wonderfully, titled Good Engineering Management is a Fad. Larson argues that what is seen as ‘good engineering management’ changes depending on the business environment. Like Chingunde, Larson started his career in the before-times, and is more circumspect about what is considered best practice:
In each of these transitions, the business environment shifted, leading to a new formulation of ideal leadership. That makes a lot of sense: of course we want leaders to fit the necessary patterns of today. Where things get weird is that in each case a morality tale was subsequently superimposed on top of the transition:
- In the 2010s, the morality tale was that it was all about empowering engineers as a fundamental good. Sure, I can get excited for that, but I don’t really believe that narrative: it happened because hiring was competitive.
- In the 2020s, the morality tale is that bureaucratic middle management have made organizations stale and inefficient. The lack of experts has crippled organizational efficiency. Once again, I can get behind that–there’s truth here–but the much larger drivers aren’t about morality, it’s about ZIRP-ending and optimism about productivity gains from AI tooling.
The conclusion here is clear: the industry will want different things from you as it evolves, and it will tell you that each of those shifts is because of some complex moral change, but it’s pretty much always about business realities changing. If you take any current morality tale as true, then you’re setting yourself up to be severely out of position when the industry shifts again in a few years, because “good leadership” is just a fad.
I feel comfortable with Chingunde’s assessment of ‘incompetent leaders’, because cheap capital makes for bad accountability. But the more important thing is that what is true for software engineering management here is more broadly true for business.
In 2018, Rahul Vohra, the CEO of email app Superhuman, wrote a famous, oft-cited essay titled “How Superhuman Built an Engine to Find Product Market Fit.” I still remember the impact the piece had on the startup noosphere — for a number of weeks, even months, it was all anyone wanted to talk about. In 2025, Grammarly acquired Superhuman for an unimpressive sum, an ignominious end for the company. With the benefit of hindsight Vohra’s essay was most likely wrong. He had not cracked product market fit. The death of his startup was also likely a function of ZIRP ending … or maybe not; who can say — this is what is difficult about all of this.
A more pragmatic approach is to modify “Everything that emerged as a best practice in the previous 15 years is now suspect …” and add “until it has been shown to give me the outcomes I desire, in this current environment.”
Again, this is where reading some business history can help. Some industries are cyclical. Think: shipping, trucking, hotels, semiconductors. In those industries it is normal to evaluate business leader performance by how well they navigate the natural capital cycles of their industries. Hot shots tend to get washed out. This is not surprising to old hands and old analysts covering such markets; it should not be surprising when it applies to other industries in the grip of longer, macro-level cycles.
It’s also not unusual to have everyone believe one thing about business, only for it to be proven wrong years or even decades later. From 1981 to 2001 Jack Welch ran General Electric, or GE. For much of the 90s he was lauded as one of the best CEOs in the world. The halo effect of his reign lasted maybe a decade after Jeff Immelt took over as successor, right as GE’s fortunes began to turn. It was gone by the time GE split into three separate companies in 2024. Welch’s performance is now regarded as suspect — remarkable for its pomp, in its time, but ultimately destructive to the firm.
Conventional wisdom that looks stupid in hindsight rarely looks stupid in the moment. Welch ran GE during a very specific market environment. In the 90s, execs who managed smooth earnings growth (against consensus earnings estimates) became rock stars in the business press. And for good reason.
Starting from about 1989, a Fortune 500 company that missed consensus earnings by a penny would lose 6 percent of its market value. GE under Welch was a master of meeting earnings estimates. The market rewarded the company handsomely. It’s a little difficult to imagine this today, but CFO magazine awarded its annual excellence awards to WorldCom’s Scott Sullivan in 1998, Enron’s Andrew Fastow in 1999, and Tyco’s Mark Swartz in 2000 for accomplishing impressively smooth earnings growth. All three ended up indicted by the federal government during the accounting scandals of the early 2000s.
Thomas King, in his wonderful history of accounting More Than a Numbers Game, calls this period the ‘EPS bubble’. EPS here means ‘Earnings Per Share’, or net income generated by the firm divided by the total number of shares outstanding.
Welch was a master of massaging these numbers, as was every other lauded CEO of the period. King writes:
The EPS bubble burst when accounting scandals came to light at the turn of the millennium. According to research from Thomson First Call, in 1998 stock prices of the 30 companies in the Dow Jones Industrial Average that beat consensus estimates by a penny saw a stock prices increase of 0.78 percent on the day of the announcement; by 2004 the effect diminished to 0.15 percent.
10 years is a long time for an effect to run. If you had a newborn when the fad started its hold on Wall Street (in 1989), your child would be 10 years old at its peak, and 15 years old for the effect to begin its decline. 15 years sounds similar, doesn’t it? It sounds like the period that we just left.
This belief in EPS growth was so strong, in fact, that there were inefficiencies in the industrials sector that Welch dominated. In Commoncog’s Case Library we covered one person — Brian Jellison — who exploited this inefficiency to generational wealth. Jellison snapped up asset light, cash generative industrial companies, something he could only do because his competitors were so focused on copying GE. He eventually turned a 100 year old manufacturing conglomerate into a large software company.
Sometimes 10 year floods also represent once-a-decade opportunities.
There is one last way in which reading biography is accretive to business skill. Our most recently published case was about Data General — a minicomputer company that rose with the minicomputer boom of the 60s, and died, like many of its contemporaries, in the personal computer (PC) boom of the 90s. It was the subject of Pulitzer Prize winning The Soul of a New Machine — one of the best non-fiction narrative books about the computer business.
A question any alert reader should ask when reading that case is “could one have saved DG in its final decade?”
If you were an operator who had cut your teeth in the computer industry, you would be hard pressed to do so. And indeed Ronald Skates, who was put in the driver’s seat at the end of DG’s life, struggled mightily before selling to EMC.
I want you to contrast Skates’s approach to that of former NASA astronaut Bill Anders, who came into General Dynamics and then did something truly extraordinary:
“In the first two years of their regime, Anders and Mellor reduced overall head count by nearly 60 percent (and corporate staff by 80 percent), relocated corporate headquarters from St. Louis to northern Virginia, instituted a formal capital approval process, and dramatically reduced investment in working capital.”
The new General Dynamics leadership, clearly, was in favor of decentralized, lean operations. As they executed a company-wide clean up, the pair discovered that most of the company’s plant managers had too much inventory on-hand. They also noticed that the managers did not focus on ROI while making additional capital requests. Mellors and Anders took decisive steps to change this. For instance, when they found an excess of expensive glass fittings for F-16 cockpits in a factory that made only one plane a week, Mellor implemented a new rule according to which each facility could only store a maximum of two extra fittings. Where he saw underutilized machinery in adjacent tank plants, he simply combined facilities.
The new General Dynamics culture strongly emphasised returns. Longtime executive Ray Lewis said: “Cash return on capital became the key metric within the company and was always on our minds.” New performance-based compensation policies were introduced, designed to reward managers for sustained improvements in stock price. Another way in which this played out was the kind of projects the company bid for. Previously, General Dynamics used to aggressively bid on most government contracts. Now, however, it only bid on projects when the returns were compelling.
All these steps drove down the working capital expenditures of the company dramatically. In Mellor’s words: “For the first couple of years we didn’t need to spend anything, we could simply run off the prior years’ buildup of inventories and capital expenditures.”
Anders recognised that the defence industry was in permanent decline, and sold off weak pieces of General Dynamics to strengthen its balance sheet. He then returned some of that cash to shareholders, and passed on a healthier company to his successors, who slowly began to expand GD into better, more lucrative businesses when the opportunities presented themselves. In this he was similar to Warren Buffett, who took over a failing textile mill and harvested its declining cash flows into a variety of other higher quality businesses. (Buffett recognised Anders’s brilliance, and became a large shareholder of GD during the latter’s tenure).
How did Anders know to do this, when Skates did not? This question is not that easy to answer. Anders had some business training at GE during the early years of Welch, but he didn’t have an MBA and was classically trained as an engineer. Skates, on the other hand, had an MBA from Harvard Business School, and was formerly a partner at major accounting firm Price Waterhouse. He worked on finance and operations as he rose through the ranks at Data General. And yet it was Skates who treated Data General as a computer business (we make computers), whereas Anders who treated General Dynamics as a business (we make, uh, money).
What I think must’ve happened is that Anders saw General Dynamics as an engineering problem to be solved, whereas Skates thought more ... conventionally. In this Anders was like precious few other CEOs. In 2012, Will Thorndike published The Outsiders, a collection of stories about ‘capital allocators’ — Anders being the subject of one of the chapters. You don’t have to work things out from first principles the way Anders did if you read the book. But if you didn’t read it … well, the odds are good that you would never work this out for yourself.
There’s a saying attributed to Naval Ravikant that goes “there’s no actual skill called business.” This is almost patently false. There are things you can do if you treat your business like a ‘system that you can engineer’. You will not learn these things if you believe the skill of business consists only of ‘marketing’ and ‘sales’ and ‘product’ and ‘hiring’ and that you must learn these things like going to the gym. But it takes a fair bit of reading to see this; it takes awhile before you realise there are businesspeople who out-class Ravikant.
I have argued that reading business history matters for the following aspects of business skill:
It’s probably no accident that businesspeople who are very good at what they do tend to read biographies (or listen, attentively, to stories) as part of their information diet. Anyone who has stuck around the game for long enough would know that business is long gestation — being able to see around corners on the order of decades is a huge edge. There’s an old saying from the racing world: “races are won in the corners, not in the straights.” Competitive advantage often comes not from doing the obvious things everyone does, in times of plenty, but from excelling at the difficult transitions.
Of course I’m biased about this. Commoncog is built around the Calibration Case Method — an approach to learning business concepts from sequences of cases. But it’s an approach to business learning that I think isn’t that obvious. After all, the aforementioned reader wouldn’t have said what he said if he understood the principle.
At the end of the day this is one of the weirder things about this game we play. There are many, many aspects of business that is like weightlifting: you can really only learn sales, management, cash flow, marketing, operations, org design, or board selection through practice.
But in the long run, knowing what has come before (and what may come after) separates the great from the merely good.
