Summary
A candid session on the key theories and practices from Joannes new book Introduction to Supply Chain. We will explore how to forecast beyond demand, move from KPIs to cash, build resilient decisions under uncertainty, and make variability a core part of daily operations.
Full transcript
Conor Doherty: This is Supply Chain Breakdown, and today we will be breaking down the major theories and themes of Joannes Vermorel’s new book, Introduction to Supply Chain. My name’s Conor—you know that—Communications Director here at Lokad. And to my left, as always, Lokad’s founder, CEO, and resident wordsmith, Joannes Vermorel. Now, manning the live chat as always: Alexey Tikhonov. Get your questions to him as soon as possible, and we will get to them a little bit later. But to the main event—Joannes, to my left, I have your new book, Introduction to Supply Chain. Room Cam 1. It’s a lovely book. Why did you write this book, sir?
Joannes Vermorel: The first few years of Lokad were rough. When I started in 2008, I started with the idea that supply chain was an extensively mature field of study and practice, with, at the time, over 60 years of literature. Now we are at 70, I would say, since the end of World War II, with a million-plus papers. I checked recently: there are over 10,000 books about supply chain in English available on Amazon. It’s a massive field, and my take was to bring this nicely packaged to the cloud with a SaaS app, while my competitors at the time—the incumbents—were still using fat clients.
Putting it on the cloud was easy. Clients did come. But nothing worked. Nothing worked, and it took me years to identify all the problems. It turned out that the mainstream supply chain theory is just focused on things that don’t work. We could even use the word “mental” for this, but it is extremely strange: you have a million-plus papers, and nothing works. You have surreal discussions with supply chain practitioners who say, “Yes, next year we are going to actually use the safety stock formula, but for now we still use something a little bit weird in our spreadsheets. Next year we will start to do the right thing; we will start to use the true serious math, and it will be good.” It turned out that’s exactly what those companies have been trying for decades.
What actually worked were classes of heuristics that are completely absent from the literature. Most of what passes for the supply chain literature just does not work. Lokad has been pioneering very different ways to do supply chain—the people who have been following this channel for a while might be aware—and I decided to have an up-to-date compilation. By the way, the theory we have developed at Lokad has been evolving. I started a world series of lectures back in 2017. It has been quite a few years. A lot of things have been refined since that time. Overall, it’s much more consistent, and there are also things that are just better.
Conor Doherty: Listening to that overview—correct me where I’m wrong—but typically when someone writes a book in supply chain, it is to either fill or address a gap. Listening to you, it sounds like you’re saying this replaces more or less all pre-existing knowledge. Or is that an overstatement?
Joannes Vermorel: It’s certainly some kind of refoundation exercise. The problem really starts with: where do you put supply chain in the tree of human knowledge? What I’m defending is that the bulk of the literature falls in two camps that are equally wrong.
Half of the literature comes in the camp of applied mathematics. The fundamental problem with this approach is that you produce supply chain papers—“theories”—that cannot ever be disproven by reality. That is very strange. Normally, if you have knowledge about something in the world and your theory is wrong, the world should be able to contradict your theory. If your theory is immune to real-world feedback, you’re doing a pure exercise in logic, in mathematics; it’s not part of the experimental sciences.
Then you have another camp—typically the sociology camp—that discusses how to slice and dice the problem across large organizations. They adopt a sociology perspective. The problem is that this perspective says nothing about the actual resolution, which is: how do you allocate your resources? How do you make decisions to govern your flow?
So, in this book I decided to have a third approach: applied economics. As surprising as it may appear, this perspective is very largely absent from the entire literature.
Conor Doherty: To switch the idea of supply chain as economics—in the book, you define it as mastery of optionality. I’m going to read this because it’s a slight update on the historical definition of supply chain: “Supply chain is mastery of optionality under variability in managing the flow of physical goods.” In plain English, how does that definition differ from the prevailing understanding of supply chain?
Joannes Vermorel: That’s the point. If you look—I can’t even really remember the exact definition given on Wikipedia for “supply chain.” The problem is that most definitions are not about “supply chain” but “supply chain management.” You already are in the realm of sociology—“I’m going to manage that; I’m going to apply some division of labor.” Most definitions are extremely broad and say all things that pertain to acquisition of raw materials, transformation, transportation, and keeping clients happy.
Most supply chain definitions take a whole paragraph and mention pretty much anything that pertains to the flow of physical goods. It’s cataloging: acquiring raw materials, transporting them, storing them, transforming them, transporting them again, servicing clients, etc. Those definitions are not crisp. If you follow them, it’s not even clear what differentiates industrial engineering and supply chain, or manufacturing and supply chain, or corporate finance and supply chain.
These definitions lack clear boundaries and an essence—clarifying a clear intent as opposed to cataloging stuff. For example, most supply chain definitions you will find on Wikipedia won’t discuss reverse logistics. As soon as you look at things that are a little bit French but still firmly in the square of supply chain, those very descriptive, enumerative definitions tend to miss them.
Conor Doherty: If you’re reimagining supply chain and the foundations from the perspective of economics, how do you square that with what I know to be a generally light touch when it comes to mathematics? Other than maybe the page numbers, you’re not going to see many integers in the book. You’ve deliberately gone for a more philosophical approach. How do people take these economic theories if it’s basically just words?
Joannes Vermorel: First, that’s a problem I have with most of academia nowadays: mathematics, as used in most papers, is filler. I’ve been trained as a mathematician—no problem there—but the math we see is not conveying powerful ideas. Supply chain is a branch of economics; mathematics is an instrument, not the point.
If I want to publish a book where mathematical instruments appear, the question is: will I convey the point in a way that is less ambiguous and more concise? That’s what mathematical formulas are about. The Maxwell equations for electromagnetism are extremely compact; in literally four equations I can convey what would take 20 pages of text. In that case, the equations bring the insights.
But when you look at the vast majority of supply chain papers using math, the math is not enlightening. The proofs are procedural and uninteresting. Give a master’s student a few hours and they will get to your proof; there are very few surprises. Even in the formalization of the problems, it’s uninteresting.
Bottom line: you end up with something tedious to read and relatively verbose, with pages of derivations that bring very little insight. In this book I decided that mathematics and algorithms are auxiliary sciences of supply chain. I try to introduce the right concepts and ideas; then people with the right background—in mathematics, statistics, algorithms—will be in a position to do the fairly mechanical derivations as needed, with the right perspective.
Conor Doherty: We’re going to get into more of the theories in detail, but a little anecdote. I want to bring you back to April last year, around the time you started writing the book—and I keep meticulous notes. It was Tuesday in April 2024. We were having a chat; you mentioned you were writing a book. I remember I asked you: Who’s the target audience of this book? I picked two names we both know—friends of the channel who have appeared here—and I said, “Is it person A or person B?” Two very different profiles, both in supply chain. Do you remember what you said to me?
Joannes Vermorel: I don’t remember.
Conor Doherty: This is not staged—you don’t actually remember. You said, “It might surprise you: neither.” Now, about 20 months later, you’ve written the book, it’s published, it’s available—who is the target audience?
Joannes Vermorel: The target audience was a little self-centered: myself 20 years ago. If I had had this book in my hands before starting Lokad, everything would have been so much simpler. It would have saved me a decade of misery. That’s a weird part of enterprise software: even if your software doesn’t work, it can be quite profitable. Not misery in the financial sense—clients are still knocking on your door—but not good.
That would have saved me and the teams at Lokad an enormous amount of time. I intentionally removed all the math from the main text—there is a little bit in the annex—because I realized it was not providing much insight. I was inspired by Basic Economics by Thomas Sowell. It’s beautifully written. Most books about economics are quite math-heavy, and I realized, reading Basic Economics—I read it to my daughter; you gave it to me as well—it’s an excellent book, excellent introduction. If you address the thing properly, you don’t need any math. The technicalities get in the way of true understanding. If it can be done for economics, it can certainly be done for supply chain. That’s the approach.
There are a few things postponed in the annex, but they are clarifying for the tech-savvy audience. Otherwise, the book is fully accessible to practitioners irrespective of their mathematical knowledge.
Conor Doherty: From the book—for anyone who has yet to read it online, it is available for free online; you can also order it on Amazon. Before we get into the weeds, one thing from the tail end: in the “Looking Onward” section you propose a concrete test for supply chain progress—“your software, whatever it is, must generate unattended, auditable (so traceable) decisions, or you must stop and explain why.” Is that, in your opinion, the ultimate goal of supply chain decision-making?
Joannes Vermorel: It’s a starting point—literally. Until you have a system capable of generating unattended decisions with what at Lokad we call 0% insanity—those decisions are at least okayish, nothing crazy—you’re not even in a position to do systematic improvement. You’re dealing with a chaotic semi-manual process, typically also quite bureaucratic. It’s impossible to benchmark anything. You can’t A/B test anything. You can’t reliably prove that whatever change you bring to the system is making it better.
Once you reach the point where you generate unattended decisions with no insanity—even if they aren’t very good yet—you have something you can A/B test. You can modify the system and dual-run: option A vs. option B, which one is best? You can have proof, select, iterate. Then the really good things can start happening: you can reason quantitatively and decide if something is making the process better or worse.
As long as you’re completely fuzzy, you have an ocean of opinions and many different people. Also, as long as you have semi-manual processes, you can have regression just because someone very experienced retired. The composition of the team changes; you changed nothing else, and you can have regression. That’s a big problem—massive confounding factors.
Conor Doherty: Obviously it’s 500 pages; we’re not going to cover everything. Over the course of the season we’ll take bits and pieces. Historically, there are overarching criticisms you have of supply chain theory. One of the most obvious—and it’s a big part of the book—is your perspective on time series forecasting, which you call a technological dead end for supply chains. Why is that?
Joannes Vermorel: That’s part of the chapter on the future. The paradigm at play is the teleological view of the future. You’re literally saying, “I can project the future and say it is this,” and, just like the plan, this description becomes the commitment. This perspective originates from the natural sciences. It’s what astronomers use to anticipate the movement of planets.
One of the early economic forecasters of the 20th century, Roger Babson, was a massive fan of Newton. His perspective—which has permeated cycles of economic forecasters, then operations research, then supply chain—was that with the right math we will soon be able to predict the future of economics, markets, everything, with the same precision as the placement of planets. It was part of progressist scientism in the first half of the 20th century.
It never worked. And you have many reasons to think it will never work. It does not preserve your own future agency. It treats the future as if it were already frozen, as if the company had no agency to change course. That is very strange.
I oppose another vision—the vision of entrepreneurs—the rugged vision, which is much more opportunistic. Fundamentally, this grand-plan teleological vision, where you treat the future as a known quantity and orchestrate everything from that, is defective. Time series are the embodiment of this perspective: they see the future as the exact symmetric of the past. You have a curve—say, temperatures in Paris, one data point per day—and you project the curve into the future. If I remove the present, nothing differentiates the past from the future; it’s the same series. If you don’t tell me where “now” is, I can’t know. It’s just a curve that goes indefinitely in both directions.
That time-series perspective, from natural science, is nonsense for business because there is an absolute asymmetry: you cannot change the past; you have agency on the future. As soon as you accept that, until you commit a resource, you have no reason to treat the future as already decided.
Conor Doherty: You go into more detail and give concrete terms: the difference between a single big buyer and many small buyers—“many small baskets versus a few large baskets”—to demonstrate the problem. Ten people buy one thing vs. one person buying ten: same time series.
Joannes Vermorel: Exactly. Time series are a very lossy representation. They compress information into a one-dimensional vector, and you lose consequential information. Imagine you have 10,000 units today for the last 10 years. What is your correct stock quantity?
Scenario one: those 10,000 units are 1,000 distinct clients; it’s very dispersed; not even the same clients every day. The odds you lose all those clients very quickly is low. Scenario two: those 10,000 units have been ordered by a single client for the last 10 years. At some point you will lose this client—bankruptcy, switch, whatever—and the potential risk is huge at all times. There is no such thing as an eternal client or an eternal company. When it happens, your remaining inventory becomes dead stock overnight with no recourse.
The time series are identical, but the risk profile is extremely different. The only way to know is to look at client composition. So, in addition to losing agency, the time series are a lossy representation of the past; you lose critical information.
Conor Doherty: How does the loss of that risk negatively impact inventory policies? This is in “The Limits of Planning.” Taking that exact example, what is the immediate-term negative outcome versus a more probabilistic approach, which you outline in the book?
Joannes Vermorel: If we step back for risk, the classical teleological view—underlying S&OP—treats risk as absent. People aren’t fools; they know risk exists, but when you look at the instruments, risk is absent. There is nothing paradigmatic about risk management. Yes, in theory you can do scenarios, but it’s second-class—an afterthought, not fitting the paradigm.
Risk is the other side of opportunity. Something unexpected can create damage; something unexpected can present an opportunity you can seize. In the teleological, S&OP perspective, the idea of seeking unknown opportunities and being at the right place, ready to address them, does not exist—just like risk does not exist.
Entrepreneurs don’t think of the future as a grand plan that is known. They see the future as fuzzy and unclear, but if you position yourself correctly and you’re prepared, you can be very lucky. I think Aristotle said that luck happens to the very prepared. It’s a different mindset and a completely different way to look at the future.
Conor Doherty: In contrast, you advance what you call the “rugged vision.” How exactly does that differ?
Joannes Vermorel: Instead of thinking you can know the future, you embrace the chaos and leverage it to your advantage. If you think of S&OP, people want forecasts as accurate as possible—in technical terms, reduce variance. What if you do the opposite—explode your variance?
In the entertainment industry you want mega hits. You’re not interested in low variance, because most attempts are failures: hit-or-miss. Reducing variance would converge to mediocrity. You want to do things so that when you have a hit, it is absolutely massive.
The rugged vision is a different take on uncertainty and variability. It sees them as a resource to be exploited, not a defect. The teleological perspective treats the future as something to be clarified, set, immobilized—your forecast and your commitment—and optimizing means achieving compliance efficiently against the plan.
With the rugged vision, it’s about exploiting uncertainty and variability. Make sure that when opportunities emerge you have first-mover advantage because you move so fast, you capture them profitably. If you make mistakes, the cost is limited. It’s okay if most bets lose as long as the cost is limited; when you win, you win big.
That creates a different perspective on what it means to plan and prepare. You embrace uncertainty and variability. It’s the same for your competitors. Instead of saying, “I’ll be more accurate,” you say, “I’ll be more reactive, agile, profitable, and opportunistic.” You acknowledge that opportunistic moves are part of the plan, and you’re comfortable not knowing exactly where you’re heading. You want to be able to respond profitably to whatever situation without needing to know exactly what it is.
From the rugged perspective, the future is radically uncertain—radically different from the past. You think about your resources as something where you want to preserve your agency and opportunities. A nice opportunity comes your way, but you might pass because you think a much bigger one is coming. You don’t want to exhaust yourself.
For example, with the tariff moves in the Trump administration, some companies anticipated distortions and brought a lot of inventory before the tariff so they could sell at the old price. The rugged vision would say: absolutely not. Now that there is this tariff, nobody can bring stuff at the same price anymore. You have no reason to liquidate. Yes, competitors will liquidate on the cheap; if it’s not perishable, that’s okay. You can hold inventory for a few months and then sell at a much higher price with a very nice margin.
Also, your agency is given by prices. That is very absent from S&OP and the teleological vision; the idea that you can play with prices as part of your plan is extremely absent.
Conor Doherty: A big part of the economics perspective is the financially adjusted or ROI-driven perspective. In the book you introduce coin-denominated objectives—you measure everything in “coins.” How is that different from the historical perspective of thinking about ROI? Is it just more all-encompassing?
Joannes Vermorel: First, in much of the academic literature—applied mathematics with an economic theme—they nominally say, “We optimize this objective function,” supposedly dollars. But that’s not a genuine economic perspective; it’s applied math wearing an economic hat. Here, we acknowledge that the economic modeling is the truly difficult challenge, not the technical calculation that results from the modeling.
Applied math says, “Give me your economic function, and I’ll do a lot of things with formulas, derive, construct theorems.” Fine. But if I take this perspective, yes, my objective function is supposed to be dollars; in practice, no one really cares. Consequently, when you look closely at many supply chain papers, what is being optimized is not even economic terms. Very frequently it’s service levels or tons of non-economic targets. That reflects the fundamentally applied-math perspective: what matters is to have a function to optimize; the nature of the function is mostly irrelevant.
Conor Doherty: To get more practical: instituting your vision—the rugged vision—required at least two roles, not just a philosophy. You introduce the roles of Flow Manager and Scientist and the importance of that pairing.
Joannes Vermorel: That is later in the book; it’s a detail. As long as, in terms of paradigm, you’re looking at the future the wrong way, you’re blocked. Nothing can happen; those things are not even thinkable. The way you organize the roles is a technicality. I describe it for clarity, but it’s of secondary importance. If you cannot even think of something, you can’t do it, no matter how you slice and dice the work or what instruments you have.
When we mention the word “planning,” it’s almost impossible to think of an alternative that is not the planning practiced despite the Gosplan in the USSR. It’s strange because the USSR was a failure, and yet large companies are literally mimicking what failed for 70 years in the Gosplan—the grand planning institute centrally piloting the entire economy.
Gosplan operated from 1925 till 1991. It produced plans for all those years; none of those plans were ever feasible. When I discuss with large companies, that’s the vibe I’ve been getting for almost two decades: grand plans that don’t work—and conceptually can’t. People think, “If we stop doing this, we stop thinking about the future,” which is not acceptable. We need to think about the future, but we cannot replace bad planning with no planning. You need an alternative. Forget the roles: as long as you cannot think properly—rugged vision as the alternative to teleology—any solution is unthinkable and thus impossible to deploy.
Conor Doherty: There are some audience questions I want to get to, but before we do—other than reading the book—assume someone has read it and is really jazzed. What are the next steps for instituting the rugged vision you described?
Joannes Vermorel: Contact us. But I think those things will come naturally. This book is an introduction, but I hope a lot of things become obvious when you have the right perspective. That’s why I didn’t go into math detail. When you know how to approach a problem, the resolution becomes procedural, technical—a mundane effort.
In school, we learn: here is a problem; identify a solution; get graded on your answer. In the real world, the most difficult thing is identifying the right problem statement. Once you can think clearly about your problem, resolving it is almost a given. In the future, I wouldn’t be surprised if, once you think clearly about your problem, you hand it to an LLM and it does the procedural thinking to give you the solution. The more intellectually demanding task is finding the right problem to solve.
Joannes Vermorel: Through this introduction, the reader can part with a very different perspective on how to even think of his company, his supply chain, and the sort of solutions that are eligible. As soon as you realize tons of techniques in supply chain are non-economic—safety stocks, for example—you won’t be surprised you don’t get return on investment. Safety stocks are not optimizing the rate of return. Don’t be surprised.
Going through the book, people will be able to say, “This line of thinking is not even in the right paradigm.” It’s a dead end; it will not generate returns because you’re not in the correct territory. Once you are, you go back to this principle at Lokad: it’s better to be approximately correct than exactly wrong. With the correct thinking and an Excel spreadsheet, you can go a long way—as opposed to being lost with the wrong take on the problem.
Conor Doherty: Fundamentally, it’s a mindset that you’re advocating.
Joannes Vermorel: Yes. For example, could the audience define, in one sentence, what economics actually is? It was defined clearly more than a century ago by Lionel Robbins, a British economist. When you ask people, usually they have no clue. I give a concise definition for supply chain, but the concise definition for economics is: “the study of scarce resources that have alternative uses.” Once you understand the words in this tight definition, you understand what economics is about.
By the way, what passes in the media for economics is not economics. It’s political ideology, or it’s economic history—“the unemployment of France is going up or down,” etc. That is descriptive; it’s not economics. Economic history requires an economic theory to make sense of it. Economics gives you that theory; they are separate concerns.
Conor Doherty: Thank you, Joannes. I’ll push on to questions from the audience. As you can see on the banner on screen, feel free to submit them privately; some comments here are DMs, but there are also public questions. This from Manuel: “This is the second book of yours. I’ve got the first one. What are the big differences between the two, or the big functional differences?”
Joannes Vermorel: The lack of math code is one enormous difference. This one is much better. The previous one was really rushed; it was done in three months. The first 100 pages are acceptable; the ones that follow are obsolete—completely outdated.
The previous book was “Here is the recipe at Lokad,” which we called the Quantitative Supply Chain. At the time, I was not completely sure about what I was pushing. I was in complete disagreement with the mainstream supply chain theory, so I said, “I’m going to do something different and call it the Quantitative Supply Chain.” But the evolution with this book is: what passes for supply chain theory in the literature is just wrong. Chapter 3 clarifies this due to epistemic concerns: where do you place supply chain—applied math, sociology, or applied economics? I argue applied economics is the correct way.
So do I have yet another theory, or do I say, “This is my best attempt at the most correct supply chain theory,” a replacement for what came before? The previous book was presented as a recipe of stuff that works at Lokad. Interestingly, that previous book is now essentially a single chapter in the new book—the chapter called “Deployment.” It covers a better, more mature version of what I described previously. The previous book is one chapter out of eleven.
I realized there was so much more under the foundations: how you see supply chain as knowledge—epistemology; economics—which I sidestepped in the previous book. We want profitability, but then you have to go back to economic roots: what does economic science bring to supply chain? The short answer is: quite a lot. Once you approach things under the explicit banner of economics—which I didn’t do previously—it clarifies tons of things.
There were plenty of other things I was taking for granted in my lectures. I realized I needed to go into the fine print, such as the modern theory of information—Shannon’s theory—which is very consequential for supply chain and how you reason about “informed” decisions. Then you have to think about the applicative landscape: software is extremely important, and supply chains are completely digitalized. I clarify the software landscape, how to think about it, and how your optimization has to be an overlay on top of it.
Then the future: teleological vs. rugged vision. That chapter came from frustration; I had hundreds of calls trying to explain probabilistic forecasting. Saying “probabilistic forecast” was the wrong approach. It’s the correct technical answer, but to understand the correct vision—independent of mathematics—you need the teleological vs. rugged lens. That gives you the underlying reasons why the instrument is desirable.
Same with “informed” decisions and intelligence. What is intelligence? With recent developments in machine learning, we can think more intelligently about intelligence. That needed treatment. Same with decisions: if supply chain is a piece of economics, every decision must be approached as an allocation of scarce resources that have alternative uses. You need to think in your company: what are the alternative uses for every resource? That clarifies a lot. There is an entire chapter dedicated to decision-making for that reason, and we will come back to it in a subsequent episode.
Conor Doherty: Next question is from Vivek—thanks for launching the book; I’m sure this will add a different perspective from all the supply chain material we have available. “Are there some use cases of real-world problems of supply chain covered in your new book?”
Joannes Vermorel: There is a place where I describe what I think about case studies—and the fact that they don’t work. But use cases are ubiquitous. Whenever you have an allocation of resources that have alternative uses, it’s a use case.
Whenever you spend a dollar to buy something, you could have used this dollar for something else—allocation of resource. If you have raw materials that can be used to produce several products and you decide to consume them to trigger a batch—that’s an allocation. If you decide finished products go to this place, they cannot be put in that place—allocation. Etc. Use cases are extremely mundane.
The supply chain literature often got this wrong because they don’t have a clear idea of what supply chain is and what a supply chain decision is. Clarify that it’s an allocation of resources with alternative uses, and the use cases become obvious: you have the flow, resources, alternative uses. Every time you choose between alternatives, it’s a supply chain decision that needs to be optimized. To know the payback you can expect, do a back-of-the-envelope calculation.
Due to shallow foundations, you end up with things that are unclear—like “supply chain digital twins.” What is it truly about? What exactly are you trying to solve? A means to what end? Many offers—consultants, vendors, professors—don’t offer a clear definition and don’t position supply chain as a branch of economics. You end up with questions like “use cases” that are hard to answer because the foundations are shallow.
This book is also an attempt to address a long-standing frustration. Implicitly, Lokad has been thinking for a decade of supply chain decisions as allocations of resources with alternative uses. But we weren’t framing it that way, and we were lost explaining why probabilistic forecast is best. It’s very relevant—but a technicality. Same with stochastic optimization—very useful, but a technicality that becomes interesting only once you understand the proper paradigm.
Conor Doherty: Next question from Nick Green—this is from our YouTube stream (we dual-stream LinkedIn and YouTube). I’ll read it verbatim: “Thank you for making the ebook free. In Southeast Asia, getting Amazon books isn’t easy. I look forward to reading. In the book, do you discuss the role of incentives in supply chain?”
Joannes Vermorel: Oh yes—absolutely. Incentives matter. You have adversarial incentives all over the place. Traditional supply chain literature largely ignores incentives and behaves like savant idiots—extremely naive with respect to incentives.
In the book I detail the incentives of employees, consultants, software vendors, and academia, and how they interplay with your attempt at doing what’s best for the long-term interest of the company. If you do not explicitly address this as an adversarial problem, it will not work. You are dealing with people who have agency; the way they act is not aligned with the company’s interest. Many situations are broken by design; you need to recognize them as such.
Importantly, with conflicts of interest, you cannot rely on the judgment of conflicted people. You can’t say, “I know I’m conflicted, but trust me, I’m honest.” That’s not how you address conflicts. When someone has a conflict of interest, this person should be barred from access to the process that generates the decision. Medical science discovered this decades ago. If you don’t do that, it won’t work.
I’m also describing the conflicts of interest that enterprise software vendors have—which includes us. I try to do a best effort here, but I am conflicted. If people want to bring things I may have omitted, post comments. I’ve been thorough about the shenanigans of enterprise software vendors—that’s something I know up close—but feedback would be awesome.
Conor Doherty: There are no other questions. Closing thought: if people only read one section of the book, what would you suggest? Obviously you want them to read all of it and buy it—but what’s one takeaway today? You talked about intelligence; you go through systems of record; there’s so much. One section?
Joannes Vermorel: Read Chapter 3, “Epistemology.” It’s the foundation of what even counts as supply chain knowledge. If you go to the end of this chapter, you will probably realize that 99% of what you’ve been reading in your life, once you understand this epistemic take, doesn’t count as part of the relevant body of knowledge for supply chain. That is strange, but that explains the “why” of my misery for the first decade at Lokad: I was fundamentally trying to use the wrong theories—like trying to get results using alchemy instead of chemistry. It’s not going to work. You’re using something that looks like science—alchemy was once very serious. Sir Isaac Newton—popular with you—spent half of his life doing research in alchemy and the other half on the movement of celestial bodies. He published two equally large books: one on celestial mechanics and the other on alchemy.
It’s not that Isaac Newton was a fool; it’s that identifying the correct paradigm is difficult. If you’re wrong, you can have something that looks like science. Don’t assume you can automatically spot what should be considered valid knowledge in supply chain. Through Chapter 3 I try to give readers intellectual tools to sort what qualifies as valid knowledge or not. Even if you disregard all the rest of the book, having this instrument will be extremely useful, because you’ll have a way to separate what is potentially correct knowledge from what is guaranteed to be irrelevant.
Conor Doherty: All right. I’m convinced. I have no further questions. Thank you very much for your time and for your answers. And to everyone else, thank you for attending. Thank you for your questions—many private comments. As I’ve said many times, some people are more comfortable commenting publicly; some prefer to message me privately. Everything that’s sent to me, I pose to Joannes verbatim, or at least as it’s given to me. On that note, be sure to connect with me on LinkedIn, and we’ll see you next week for another live episode of Breakdown. And on that note—get back to work. Check this out.
