The Straitjacket of Legibility

In 2025 we ran our first cohort-based course, titled Speedrunning the Idea Maze (StIM). It was our very first live course, and it was designed to test two hypotheses — or at least, two that I’m comfortable sharing in public.

The first hypothesis was that we could amortise the cost of case production across a higher-priced product. One may charge higher prices for live courses, first because the experience is perceived as more valuable, and second because more can be done with the medium (it is difficult to transform students through writing alone; it is more possible if you’re designing a series of live exercises with feedback). We were already paying the costs of case production anyway, so what happens if we can repackage it where it brings in more revenue — that also results in more value for our customers?

The second hypothesis was that we could accelerate our participants’s understanding of the idea maze, by compressing decades of experience into a few hours. The basis of this belief, and what we learnt executing it I have covered elsewhere.

What I want to talk about today is one very specific thing that we covered whilst executing the second cohort of this course — recently concluded as of 13th February 2026. I want to talk about the reaction our students had to one idea, and then use that as a jumping off point to talk about the idea in business more generally.

First, a quick summary of the primary concept of the course.

StIM’s core concept is Effectuation — that is, the observation, originally published in 2001 by academic Dr Saras Sarasvathy, that all expert entrepreneurs think in the exact same way. In StIM, our central challenge was to get students to see effectuation in the wild and in their own lives. In parallel, we got students to think effectually in response to hypothetical scenarios.

Our end goal was to change the way they saw things. Successfully changing perception changes behaviour: if you are able to think effectually you will see a range of possible actions you’d never consider. As a result you will act more entrepreneurially.

The word ‘effectuation’ comes from ‘effect’ in ‘cause and effect’. The easiest way I know to describe the idea is to talk about two approaches to cooking for a dinner party:

1. In scenario one you are holding a pizza party for a group of five friends. You plan out your menu, figure out what it takes to make enough pizzas for your friends, and then you go to the supermarket to buy those ingredients and prep dough and toppings. Sarasvathy calls this causal thinking — you are setting a goal, working backwards into a plan, and then executing that plan.
2. In scenario two you are holding a dinner party, and you don’t really know what you’re going to make. You open your fridge to figure out what ingredients you have. Based on those ingredients, you improvise your way to a dinner menu. This is effectual thinking — the endpoint is indeterminate and you are working based on what you currently have (and also what you can easily create).

The point that Sarasvathy makes isn’t that one style of thinking is better than the other — entrepreneurs do a mix of both — it is simply that when starting a new venture, entrepreneurs think more effectually than causally. In fact, the set of actions that you do as a result of effectual thinking is what lay observers would call “acting entrepreneurially”. They are materially different from the kind of causal thinking you would observe in more corporate environments.

The actual process of effectuation in entrepreneurship proceeds as follows: all good entrepreneurs start out from 1) who they are, 2) what they know, 3) whom they know and 4) what they have, and effectuate forwards into the unknown, in search of some customer demand they can serve … in a configuration that is acceptable to them.

As they do so, they follow the following three principles:

1. They take many affordable loss bets, using each bet to generate new information.
2. They have an uncanny ability to turn stakeholders into partners. (Partners here are defined as parties who have a vested interest in their success). Most typically these are their customers, but many entrepreneurs are also able to create situations where their investors, suppliers and even landlords have a stake in them and want them to succeed.
3. They are ok with many of the outcomes they are able to effect. So for instance they are comfortable iterating forwards and going “Oh, I guess the only viable business configuration for this problem is a consultancy. Alright then!” And instead of looking at it as a dead end, they tend to treat it as a means to build more resources from which to effectuate forwards from, in the near future. (Or maybe not, they may also just choose to give up and try again).

You see how this might show up in actual stories of entrepreneurship. In business school, new venture creation is often taught through the logical method of figuring out a market to sell to, picking a product to make, coming up with a market positioning, calculating the total addressable market for that positioning, thinking about potential competitive advantages, and then raising funding to attack the identified opportunity. This is causal thinking in action, and it is what the academy does best. Of course you’ll want to work backwards from some set of goals. Of course you’ll want to create a business plan. And of course what is taught bears no resemblance to how entrepreneurship is done in the real world.

In practice, entrepreneurial stories tend to be random and highly path-dependent. Less “we planned this” and more “We got our start making engines at old Otto’s gas-engine works over in Cologne, and then Gottlieb thought he’d like to make one small enough to bolt onto a horse carriage and he went off and did exactly that, and then we thought, ‘hmm, what if we got into making proper motor cars?’ And before we knew it we were selling cars and folks really seemed to like them! Then we expanded to France because this Austrian fellow Jellinek over in Nice wanted to race our cars on the Riviera and asked us to name them after his daughter Mercedes, and it seemed so easy to just set up a sales office …”

(Yes, that is the actual origin story of Mercedes Benz).

Some interesting things fall out of this framework, once you sit and think about it for a bit. In StIM, we pointed out that one consequence of effectual thinking is illegibility. That is: since you’re improvising in response to demand, settling only for configurations that are acceptable to your unique requirements, you’ll often find yourself in weird situations where things are working but you have no idea how to describe what you do. Mercedes Benz is a coherent car company today, but it was an oddball operation during its early years. In many cases the entrepreneurs themselves don’t know why what they’re doing works, only that it seems to. Figuring out the ‘why’ can lead to growth, but this initial insight may remain under-publicised for many years. Over time, though, success leads to attention, which leads to analysis, which ultimately leads to more legible explanations. But it is the success that leads to legibility, not the other way around. In other words, when you are effectuating, the odds are pretty good that you’ll end up in a highly illegible place. You should expect to have great difficulty explaining what it is you do to your friends and family.

When we presented this idea to our cohort, folks started nodding their heads. “I feel this so much,” one student said. Another student, who was more experienced with entrepreneurship, said: “I’ve been illegible my whole life. In my current venture our partners want me to make it legible right now, but I feel like that’s entirely the wrong move.”

In class, we pointed out that novice entrepreneurs tend to go after legible things too early. They often sought legibility before they were sure that the business they had built actually worked. Why do they do this? We posited that they did this because they cared about identity, and that they cared about what others thought of them. In my experience, such entrepreneurs want to be able to tell their friends and family “we make AI-powered contract software for lawyers.” They want to go on conference panels and be addressed as “John Doe, CEO of ABC Corp, the enterprise platform for AI-powered customer support AI agents.”

But this is a mistake. Good businesspeople are perfectly willing to sacrifice legibility in the search of a unique configuration that works for them. One thing that I like to tell folks is that the late great Charlie Munger was completely illegible for a period between his 30s and 40s. From Damn Right, journalist Janet Lowe’s biography of the man (all bold emphasis mine):

While Munger was going through this period of enormous change, expansion, and development, his family sometimes were in a quandary when they tried to describe what Charlie did for a living. Molly Munger had trouble explaining her father to her friend Alice Ballard, a Philadelphia blue-blood and debutante. Alice had scored 800 on the verbal portion of the SAT and like Molly, attended Harvard Law School. To a California girl, Alice seemed worth impressing, but Charlie was of no help.

“My college roommate’s father was a partner in an old-line Philadelphia firm. She was descended from William Penn. Charlie called on Fred Ballard (Alice’s father), who later said, ‘I have no idea who [Munger] is from what he said of himself. He could be working for the CIA.’ Daddy made no coherent explanation of himself. He had this ratty little office in the stock exchange. What did he do — working in a ratty little office? And a fledgling law firm … he bought odd companies like K&W — the automotive chemicals company.” Nevertheless, Molly’s faith in her father was unshaken. “It didn’t matter to me. It seemed like, ‘You just don’t know him. If you don’t know it now, you’ll know it later. He’s fabulous.’”

At the point of his death, Munger was known (and widely celebrated) as a legendary investor, the vice chairman of Berkshire Hathaway, and the right hand man of Warren Buffett. But that was much later. In their 30s through their 40s, Buffett and Munger were highly illegible. Their business dealings were so odd, in fact, that the SEC opened an investigation into their cross-holdings in 1974. This forced the pair to consolidate their activities under Berkshire Hathaway, and to take up the role of Chairman and Vice Chairman respectively. Oddly enough this move made them more legible; this legibility persisted to the end of their careers.

I want to pause here and draw your attention to what it must’ve felt like. In the 1970s, Munger was a) running an investment partnership, b) had started a law firm with his old colleagues, c) was wrapping up the last of his real estate development projects in LA, and was d) doing random business takeovers with one Rick Guerin and a folksy 30-something oddball named Warren, based out of Omaha. I want you to put yourself in Munger’s shoes. If someone asked you what you were doing during this period … what would you say?

A more contemporary example might also help. I still remember when TikTok burst on the scene in the late 2010s. I remember feeling confused. My frame of reference was Facebook, Twitter and Instagram: social networks, where you had to ‘friend’ or subscribe to see a person’s posts in one’s feed. I thought that TikTok would also be a social network, but everything about the app felt off. Subscribing to folks didn’t seem to make much of a difference to the content one saw; there was no feed, not as I understood it. And it was clear that it was very, very addictive. It was much later that I (and collectively, the entire tech industry, probably) understood that TikTok was some other, more interesting thing.

TikTok was created by ByteDance, and ByteDance founder Zhang Yiming is a pretty good example of what we’re talking about. ByteDance started out making throwaway apps. They saw themselves as a machine learning company, not a specific app company. Their first success was Toutiao — a news app, powered by a recommendations engine they then spent years perfecting. This same technology was later applied to short videos. Zhang wasn’t thinking about existing app categories when he created TikTok. He was a good businessperson. He cared only for what worked; he let the categorisations come later. Today, of course, TikTok is legible. We understand better the strengths and weaknesses of its model. But it was illegible for a long time, and often misunderstood. Zhang Yiming didn’t care.

An Advantage

In business, illegibility is often an advantage. In last year’s letter, I wrote that the ‘biggest problem in business’ is competitive arbitrage: you make something good, then people notice, then they copy you and cut their prices, then others copy them and cut their prices, and then everyone in the category spirals downwards in price competition until margins bottom out at the opportunity cost of capital.

Illegibility is often a good first defence against this. Nobody can copy you if they don’t understand what you’re doing. This is not new; the venture investor Jerry Neumann is known for saying that “uncertainty is every startup’s first moat.” Without uncertainty, large enterprises will instantly spot the full consequences of a new business, and will act to crush startups at infancy. (We are seeing this at play out in the current AI moment, for instance.)

Uncertainty isn’t exactly the same thing as illegibility, though both are related. Uncertain things are illegible, but not every illegible thing is uncertain. Illegibility may come from many other sources. It may come from doing an old thing in a new way. It may also come from perceived incoherency — that is, multiple things that may make sense that do not appear to make sense when put together. (I write perceived incoherency because it is important to have actual coherency — otherwise it is not likely that what you’re doing actually works).

I don’t want to overstate this and say that illegibility is something to aim for. It’s more accurate to think of illegibility as the most common side effect of a good process. To understand why, it’s instructive to invert the question: what is legibility, really? Legibility occurs when what you are doing fits into one of a few known forms in the heads of others. Those forms exist because they are well-worn, and therefore well understood. A doctor is legible, as is an M&A lawyer. On the other hand, a 40 year old who invests in equities through an investment partnership, owns control investments in multiple holding companies, issues warrants from some of those companies, acts as a founding partner in a law firm and does real estate developments on the side is a shape that fits no known form.

So on one hand you have legible forms. On the other hand you have unique individuals. As a unique individual, it is highly unlikely that everything that you want in your life fits a known form. If you wish to find a unique configuration that fits the contours of all you want, it’s highly likely that you’ll have to break the straitjacket of your chosen form. This leads to some illegibility.

(The alternative, of course, is to change your preferences, to fold down your dreams and to set them aside).

But ultimately, fitting what you do to what you want leads to relatively mild illegibility. The strong form of illegibility — the thing that causes the organisations you build to be truly weird, is when you have a willingness to break known forms in the pursuit of what works. A business mentor and I were talking about this in passing last year. His quip was simply “if it works, it likely doesn’t make sense to the average outsider.” This applies broadly.

In fact, “does it work for your chosen goals?” is a good test for whether the illegibility you’re looking at is the result of a good process. Some folks are illegible because they lead fundamentally incoherent lives — what they are doing does not get them closer to what they actually want. Such illegibility is not the outcome of a good process.

To put this a final way: if your goal is to do what works, it is almost guaranteed that there is no clean label for what you eventually figure out. The high order bit is that it works and it fits your goals, not that you can put a name to it.

As with many such things, this is a tradeoff. But I think it’s worth it.

Cedric Chin
27 April 2026