00:00:00 Introduction to the interview
00:02:08 S&OP book and debate
00:08:01 S&OP as a business process
00:13:25 Strategic decisions in aviation
00:18:12 S&OP’s link to company maturity
00:23:51 Strategic briefs and Amazon’s success
00:30:23 S&OP decision dilution and ownership
00:37:23 Software-centric information flow in S&OP
00:44:13 Executive reliance on system advice
00:50:11 S&OP meeting cycles and efficiency
00:55:33 Market share strategy and decision-making
01:01:46 S&OP as a journey of maturity
01:07:04 Costs and benefits of S&OP implementation
01:13:15 Overstaffing toxicity and lean teams
01:19:04 Microsoft’s S&OP and innovation
01:24:47 S&OP’s rigidity versus flexibility
01:30:22 S&OP’s future and technology trajectory
01:36:36 Job evolution and digitalization’s impact
01:43:28 Challenging traditional operations with S&OP

Summary

In a recent LokadTV interview, Conor Doherty hosted a discussion between Joannes Vermorel, CEO of Lokad, and Eric Wilson, Director of Thought Leadership at the Institute of Business Forecasting (IBF), focusing on Sales & Operations Planning (S&OP). Wilson, advocating for S&OP, highlighted its role in aligning company functions for coordinated decision-making and its benefits for mature companies. Vermorel critiqued S&OP as outdated, advocating for a software-centric approach to enhance efficiency and reduce human intervention. The debate covered the effectiveness of meetings, incremental versus radical changes, opportunity costs, and the future of S&OP, with both agreeing on the need for significant technological advancements.

Extended Summary

In a recent episode of LokadTV, Conor Doherty, Head of Communication at Lokad, hosted a conversation between Joannes Vermorel, CEO and founder of Lokad, and Eric Wilson, Director of Thought Leadership at the Institute of Business Forecasting (IBF). The discussion centered around Sales & Operations Planning (S&OP), its role in modern supply chain management, and its long-term viability as a business practice. This conversation served as a follow-up to a previous debate on the same topic.

Eric Wilson, with decades of experience in supply chain and S&OP, shared insights from his career and his recent book, “A Practical Guide to Sales and Operations Planning.” He emphasized that S&OP is a business process designed to align all functional areas of a company around a unified set of assumptions for coordinated decision-making. He argued that while many companies are still in the early stages of S&OP maturity, those that reach higher levels of maturity see significant benefits, including improved bottom lines and reduced uncertainty.

Joannes Vermorel, on the other hand, critiqued S&OP as an inefficient and outdated process. He argued that the traditional S&OP process relies too heavily on human-mediated information flow, which he sees as obsolete in the digital age. Vermorel advocated for a more software-centric approach, where economic variables are embedded into decision-making processes, allowing for real-time adjustments without the need for frequent meetings. He believes that this approach can significantly reduce the need for human intervention and increase efficiency.

The conversation touched on several key points of contention:

  1. Role of Meetings in Decision-Making: Vermorel argued that large meetings often dilute responsibility and do not contribute meaningfully to strategic decision-making. He cited examples of companies where strategic decisions were made by a few key individuals rather than through committee consensus. Wilson countered that structured S&OP processes, when done correctly, can enhance accountability and communication within an organization.

  2. Incremental vs. Radical Change: Vermorel expressed skepticism about incremental improvements, arguing that true breakthroughs come from radical changes. He cited examples of companies that have successfully automated large parts of their supply chain management, resulting in significant workforce reductions and increased efficiency. Wilson acknowledged the potential for automation but emphasized that many companies are not yet ready for such a leap and need to go through a journey of incremental improvements.

  3. Opportunity Cost of S&OP: Vermorel highlighted the opportunity cost of maintaining traditional S&OP processes, arguing that they prevent companies from pursuing more innovative, software-driven approaches. Wilson responded by pointing out the measurable benefits that mature S&OP processes can bring to companies, including improved financial performance and operational efficiency.

  4. Future of S&OP: Both speakers agreed that the future of S&OP will involve significant changes, driven by advancements in technology and automation. Wilson suggested that future S&OP processes will focus more on business efficiency and strategic decision-making, while Vermorel emphasized the need to simplify and streamline processes to achieve greater efficiency.

In conclusion, the conversation highlighted the ongoing debate about the role and effectiveness of S&OP in modern supply chain management. While Wilson defended the value of structured S&OP processes, Vermorel argued for a more radical, software-driven approach. Both agreed that the future of S&OP will involve significant changes, but they differed on the best path to achieve those changes.

Full Transcript

Conor Doherty: Welcome back to LokadTV. Today, I have the pleasure of hosting a conversation between Eric Wilson and Joannes Vermorel. Eric is the Director of Thought Leadership at the Institute of Business Forecasting (IBF) and the host of the IBF On Demand podcast. Eric is a vocal proponent of S&OP, and this conversation is intended to serve as a nuanced follow-up to the debate Joannes Vermorel had with Milos Vrzic several months ago. As always, if you like what you hear, subscribe to the YouTube channel and follow us on LinkedIn. With that out of the way, I give you today’s conversation between Eric Wilson and Joannes Vermorel.

Well, first of all, Eric, thank you very much for joining us. You are a man who wears many hats. You’re a writer, a podcaster, a director of thought leadership, a consultant, and a public speaker. Those are a lot of titles. Could you perhaps, for our audience who might not be as familiar, give more insight into who is the man behind those titles?

Eric Wilson: Well, the man behind the titles actually has a few decades of practitioner experience. I have scars to prove it. I did a lot of transformational projects with many organizations, worked in supply chain, S&OP, and wore all the different hats there as well. Now, I’m currently the Director of Thought Leadership with IBF. I do some consulting as well, which is just a fancy way of saying that I liaise between what companies are doing now and share it with other organizations. I work with organizations, do in-house training, conferences, writing, and facilitate discussions like you do with organizations about what we see in the industry and where we see it going.

Conor Doherty: Well, thank you. And again, mentioning writing, that was actually more or less how we first got in contact about this discussion. You mentioned that you had written a book, “A Practical Guide to Sales and Operations Planning.” You sent us over a copy, thank you very much. I’m just curious, did you pour all your scars into that book? What was the journey like?

Eric Wilson: Yeah, it was actually my second professional book. I had “Predictive Analytics for Business Forecasting” about three years ago. My most recent book, “A Practical Guide to Sales and Operations Planning,” came about because there really is a void when it comes to a modernized text in the data and AI era, post-COVID, of what that process is. So, when I say I poured my scars, yeah, I did, but I also poured everybody else’s scars into that as well.

The beauty of that book is I didn’t write 40% of it. I had a co-author who helped me write a lot of the book. A lot of that book was curated from the IBF, which has been around for 42 years. It’s a professional organization built by practitioners for practitioners. We have a quarterly Journal of Business Forecasting with great practitioner articles, peer-reviewed articles. About 40% of this book is curated from those articles, with other subject matter experts and thought leaders in the field. We synthesized that information to have a one-source that shows where S&OP is today, where it’s going, and the nuances of it from multiple perspectives. That’s what makes it special.

Conor Doherty: Well, actually, that’s a really salient point to raise because one of the reasons we wanted to bring you on in particular was that Joannes, several months ago, had done a debate on S&OP with Milos Vrzic, a friend of the channel. We wanted to revisit some of the claims that were made in a less formal, less timed, and rigorous structure. Reaching out to you meant also getting access to all those perspectives you just described. So, one of the reasons I’m looking forward to this is it’s not just your perspective, but having curated that book, you can cite the experiences of many other practitioners.

On that point, let’s get into it. As I said in my introduction, the idea here is to have a more nuanced, freer follow-up to the debate that Joannes had done. What I wanted to do was take some of the key claims that Joannes had made, some provocative, unlike your style but a little bit provocative.

Joannes Vermorel: It may happen accidentally, you know.

Conor Doherty: It can happen. And then, just go back and forth and see where we agree or disagree, rather than just a general ranging discussion where everyone jumps off into different directions. So, before we get into that, I’ll ask you both, starting with Joannes, for your executive summary when it comes to S&OP. What is it that you dislike? What is it you like? What is your position here today about S&OP?

Joannes Vermorel: So, first, the way I think of S&OP is I think about the problem. The problem is large companies that operate large supply chains have a big synchronization problem. It means that you have thousands, potentially tens of thousands of people doing different things at the same time, and ultimately, you want to orchestrate all of that so that you serve your clients better and more profitably.

This synchronization, the fact that what you’re promoting, what you’re going to sell, that you manufacture what you’re going to sell, that you have in stock what people want to buy, all of that is not a given. You need a way to achieve this kind of company-wide synchronization between demand and supply and many other forces. So, that’s the problem, and I see S&OP as one solution, a candidate solution for this problem.

Conor Doherty: So, you like it?

Joannes Vermorel: No, I say it’s one solution. I don’t say it’s the best solution. I think it’s not a good solution. In terms of characterization, I think it’s a solution that emerged from a paradigm that is very much people-centric. It means that information flows through people, and the idea is to have a series of meetings so that we will have this information flowing through the company. And then, through this process, as it’s very paper-oriented, we will achieve the synchronization.

I see that as the biggest flaw, as opposed to an alternative paradigm that would say machine-first, where information flows from computers and people are only involved in structuring the flow of information and structuring where decisions are computed, as opposed to being involved with the flow and being involved with the decision. That would be the crux of my position.

Conor Doherty: Okay, and Eric, you’ve heard a lot. Again, we’ll get into the details of some of those claims a little bit later, but you’ve heard Joannes’ overview. Your thoughts?

Eric Wilson: What I just heard was supply chain management. I think that’s what you described. The problem is supply chain management. It’s complex, and we try to oversimplify it. As soon as you have the people aspect, you’re going to have limitations. We’re all about flow, and when you go textbook supply chain management, it is goods flow and information flow. When you add people to try to manage that, you have issues.

My perspective, and what I picked up when I looked at S&OP more holistically, is that sales and operations planning, IBP, I don’t care what we call it, it’s all the same. It’s a business process that aligns all functional areas around a unified set of assumptions for coordinated decision-making. The key is that it is a business process. It’s not supply chain management; it’s a business process. We’re aligning all functional areas.

It’s not balancing supply and demand; we’re aligning the company. We’re aligning all functional areas around a unified set of assumptions. I’m not saying one number, but we’re aligned to the assumptions and the strategy of the organization. That’s the key, and it’s for coordinated decision-making. It’s a decision-making forum. It’s not transactional; it’s not trying to create a plan for tactical operational standpoints. It is a decision-making forum for business. That’s what it really is.

Conor Doherty: Joannes, does any of that resonate with you?

Joannes Vermorel: I see your position. My take is that the distinction between supply chain and business, for me, the frontiers are very fuzzy. Again, I prefer to go for things that are a little bit more factual, such as counting the decisions. Let’s survey the decisions. For every product, do I invest, and how much, in advertising and where? What will be my price point? Do I buy materials to produce them? Where do I produce them, in which quantity, according to which timeline? We can enumerate all those decisions.

Some decisions belong to what people would call supply and demand planning, like purchasing materials. Some, like pricing, usually do not belong to this area; they belong to other areas, typically marketing. The budget investments, for example, budgeting how much advertising is going to be spent, that’s another thing. My take is that at scale, when we talk about those decisions, we are talking about large-scale decisions. A large-scale company has tens of thousands of daily decisions, potentially millions if we’re talking about a $10 billion-plus company.

My take is that when people describe a business process to generate those decisions, I would say yes, obviously, those decisions are made. But the challenge is, does the business process you describe contribute to making better decisions? My counter-proposal to that would be those committees, because that’s what it looks like whenever I see an S&OP process in action, it is like a series of meetings that look like committees. Those committees do not contribute meaningfully to improving those decision-making processes when it comes to generating tens of thousands of daily decisions.

Eric Wilson: You’re looking at those thousands or, you know, million different decisions that are done daily. I fully agree. And when you’re looking at automating, you have the capability of automating those decisions because everything you talked about was transactional pricing, you know, where the inventory should be, manufacturing decisions. Those are transactional decisions, and I 100% agree a lot of those should be automated. You can create efficiencies and better precision with some automation there.

What you aren’t able to automate right now is the strategy side of things. When you’re looking at a product committee, product review, the output of that is not “here’s the SKU I’m going to do” or “here’s the SKU I’m going to discontinue.” The output of that is a brand strategy. The brand strategy drives the systems, drives the people. Right now, as far as the tactical execution and transactional decisions coming out of a demand review, the output’s not a forecast. The output is “what is my channel strategy?” “What is my customer strategy?” That is the output of a demand review or that demand committee.

The transactional decisions of that are “what are we going to order?” “What are we going to do?” “Where are we going to place the information?” “How are we going to price it?” Those are the transactional outputs. Same with supply. Supply ultimately isn’t just balancing and transacting; it’s an optimization or fulfillment strategy, and that’s based on what’s going to be right not only for the customer but also for the company. When you get to the strategy side of things, that’s what really drives a mature S&OP process that is beyond transactional, and that’s the part you’re not automating yet.

Joannes Vermorel: I very much agree on the fact that strategy is not part of the realm of stuff that we can automate yet. I agree with that. Now, my counterargument would be I’ve never seen a meeting with more than, let’s say, 10 people do anything of strategic relevance. You know, and you see the problem. The problem is that most of the strategic decisions are terrifying. They are brutal. People are going to be turned to dust through the process.

Just to give you an example, let’s imagine we have a strategic meeting with Boeing. What would it entail? It would be a discussion about what if we have two planes that crash next week, killing 500 people. And what next? What do we do? And that’s a very real possibility. I’m not saying that it’s likely; I really pray that it doesn’t happen. But is there every single week a 0.1% chance that it can happen? Yes.

Now, it’s not much, but it’s not zero. And so you see, I’ve never seen any of those life-and-death, brutal discussions happen in a kind of committee. Sometimes you need to go for very difficult decisions. For example, Boeing has been in a series of issues. I’m taking on Boeing just for continuity of the example. Boeing had been sleeping in terms of engineering culture for two decades. I don’t know if it’s SpaceX that just happened to hire all the good engineers of Boeing or whatever, but that’s a very tough problem that needs to be addressed.

So, your strategy, that’s where I said I do not see strategic decisions ever being made in this sort of company. If we look at another historical example, it would be Nokia, who went from paper production because Nokia was doing paper, then they did phones, and then they lost it all because they never transitioned to smartphones against Apple. But you see, those sorts of strategic inflections, I do not see that ever happening in a committee.

In the committee, the typical thing is bike-shedding. I’m not sure if you’re familiar with this concept of bike-shedding. That’s from the laws of Parkinson, where people take an easy issue and they will bike-shed like, “Oh, this super stable product category, will it grow 3% next year or will it shrink 2% next year?” Those are gentle questions; they don’t challenge anything. I don’t see in a committee ever people discussing things like, “You guys, all of you should just be fired.”

It’s very… Imagine again, just look at Elon Musk, what he did at Twitter. Over the course of one year and a half, he fired 90% of his company. Some left, some were fired. This is not the sort of thing that is very strategic, but can you realistically have a discussion in a meeting with 20 people where you’re going to discuss, “We are going to fire half of us. Let’s vote.” I mean, sorry, it’s just…

Conor Doherty: Well, if I can just build on that point because I also do want to demonstrate that I read your book, Eric. I went straight to the table of contents because I knew I had read about what is efficient decision-making. So, you have a section there in your book, there are six elements of what you… I know there’s an essay later written by somebody else, but I’m presuming that the unlabeled sections are from you. And there is one section on efficient decision-making where you discuss scenario planning. And I’m going to quote back, and then the question comes.

So, given the inherent uncertainty, as Joannes mentioned in his Boeing example, the inherent uncertainty in business environments, S&OP decision-making often involves scenario planning. Planners develop multiple scenarios based on different assumptions about demand, supply chain disruptions, planes falling out of the sky—you didn’t write that—and other variables. By simulating these scenarios and assessing their potential impact, decision-makers can evaluate various courses of action and develop robust contingency plans to mitigate risks.

So, having read that, and it sounds cool, I’m all on board. Coming back to the example that Joannes just gave of, okay, if you’re at an enormous company like Boeing, there are low probability but potentially catastrophic events that could happen. Are you going to stand up in an S&OP meeting and say, “All right, well, if one plane falls out of the sky, here is our financial risk?” I mean, that’s an unpopular and very tricky thing to cover in real-time, face-to-face with many people. So again, my question is, do you see those kinds of scenarios, low probability but do happen, being covered in the types of processes and S&OP meetings you advocate?

Eric Wilson: It depends. It depends on the company, it depends on what they need to focus on, what they should focus on, but it also greatly depends on the maturity of the organization and the maturity of the process as well. Because even the statements that you just read from my book, when you’re looking at scenario planning, that in itself has a maturity inside of an organization. And half of the organizations, probably close to 60%, two-thirds of the organizations out there fit within the exact definition that you were describing, saying that they focus on the wrong things. They’re doing scenario planning because they don’t have the ability to enable data and technology to do it for them.

So, they have this as a resource and a process to help enable those decision-makings for that level of maturity in that organization. And for them, absolutely, they’re talking about supply and demand balancing, they’re talking about what if we do pricing, what if we do… and they’re talking about exactly those things that you mentioned we could automate, but they’re not. They have this as their resource to be able to make those decisions.

As organizations continue to get more mature, we have seen organizations out there, and they do exist, that do have more strategic decisions. Target is a great example. Target looks at being able to plan their footprint long-term based on decisions of where they want to go to market, cash, and a lot of other executive strategic decisions that are now being compiled and saying, “Hey, based on this, this is what we want to do.” We see the same things with Microsoft. These organizations have more mature S&OP processes that are bringing strategy now into their organization. So, even the what-if scenarios, that becomes a maturity, and it takes on a little bit different meaning as you get more mature.

Conor Doherty: Thank you, Eric. And I do want to come back to get your take on this, but I want to add a little bit because a term you used there, Eric, maturity. So, companies to have more mature processes. So, Joannes, in your answer, do you agree with the idea that to make the kinds of pivots in terms of better decision-making that you would advocate, that maturity is an aspect? Because I know you previously had made a nice distinction. You used words I’d never heard before: precocial and altricial. So, some animals are born precocial, meaning…

Eric Wilson: I don’t even know those words exactly. It’s a big word.

Conor Doherty: He speaks English better than I do. It’s ridiculous. Precocial animals are born and are able to walk almost immediately. So, certain animals like horses, for example, they’re born, they can run and frolic and all of that immediately. Altricial, humans, there’s an enormous maturity phase for humans. You’re born and you don’t walk until what, two, three? Anyway…

Eric Wilson: That sounds like most S&OP processes.

Conor Doherty: There, so that’s, and that is my point. So again, we don’t need to use the words precocial and altricial but are you on board with the idea that the kind of pivot you would advocate is predicated on any kind of maturity? Or is it something that can just be…

Joannes Vermorel: I’m very business-wise, I’m extremely skeptical of anything that involves incrementalism. That’s a term that I use with a pejorative tone. Most of the pivots are always a giant leap into the unknown. You do not incrementally become like your target. You do not incrementally become an e-commerce player. Either you do, or you don’t. Nokia did not incrementally become a phone manufacturer. They were doing paper, and at some point, they were doing smartphones.

If you think that there is a path where step by step they did that, it is just… same thing with even Apple. They didn’t become incrementally a phone manufacturer either. I think that that’s a very dangerous mistake because it gives this sort of fallacy of “oh, you have to learn to walk before running” and whatnot. That works for humans and sports, but it does not work for businesses.

And if we go back, for example, to your strategy, so I was looking at contingencies not being discussed. That was an example of that. But what would be a very endgame sort of strategic production that would represent the highest sort of maturity? One public and super well-known example is a document, a 20-page document called “A Testament of a Furniture Dealer,” which was written in 1970 by Ingvar Kamprad, the founder of IKEA. This document is a 20-page strategic brief that 50 years later is still of prime relevance. I read it; it’s incredible. This thing is incredibly visionary. Fifty years later, it is still incredibly relevant for IKEA. Wow.

And the magic is that half of it is even relevant for my software company, which has nothing to do with IKEA. So, I mean, it’s a testament to how good the document is if half of it is relevant for a company that is in no shape or form even closely relevant to IKEA. But for IKEA, it is still, 50 years onward, completely relevant now. And this is beautifully written, extremely insightful, and short and concise. Again, I have never seen a committee produce these sorts of things. I have seen over and over people producing strategic briefs of very high quality. This one is like a historical landmark in a way.

There was plenty, for example, Jeff Bezos sending a memo in 2002 saying that everybody had like two weeks to have a plan to go for API first, or the managers would be fired. That is also this sort of super important strategic brief that explains a lot of the success of Amazon that followed. But my take is that, again, I’ve never seen especially mature companies produce these sorts of super high-quality documents. Usually, they produce insipid soup. They produce a document that is what I call corporate happy talk, where you just make everybody happy. Every sentence you say, “Oh, we should consider that, but we should also consider the opposite.” It will only be like super mild temperature for everything. There will be no radical conclusion. Nobody will ever be threatened to be fired in one of those strategic memos. There will not be super direct blame assignment, saying this person is to blame, this person needs to be fired, and maybe with litigation as a double down because there was literally fraudulent negligence. And yet, sometimes it is literally the most honest strategic diagnosis there is.

So you see, my take is again, because when I look at S&OP the deliverables produced by those committees, because that’s what those meetings look like, you know, 10 people or more, do they produce deliverables that are really, really good? And sometimes to the point of becoming historical landmarks. And I really mean it. Do you have a chance out of those committees to produce a document that is so good that 50 years from now historians will look at it and say, “This was a time where this company made the real strategic inflection that just ended up in doing one of their corporate values”? Just like, let’s say, Apple before iPhone, after iPhone.

Conor Doherty: Eric, I’ll come to you in a moment, but I just want to clarify. Are you identifying, just so I understand and presumably the audience as well, are you saying that if a given meeting doesn’t yield a new Bible or new verse, it’s a waste of time?

Joannes Vermorel: I say if this class of meetings never does, it is a problem. If it never does on average. Obviously, I’m pretty sure that on average, the average memo produced by Steve Jobs was so-so, you know, batting a thousand. So I’m not referring to the average. I’m just saying that if you want to look if a process can produce greatness, you would look at least at the sort of… do you have an example of greatness to be found in the past? I can find tons. Yes, I can find literally tons of great strategic memos. Especially, I love Silicon Valley. People are very open; the memos are leaked, probably intentionally. But I would say once a week, there is a company from the US where there is a strategic memo that is leaked, and I think, “Damn, this is good. This is really good.” I wish I could produce this sort of super keen, concise, well-written strategic insights and bring all my company on board.

I have attended dozens of S&OP processes, which turns up probably over 100 meetings. I have never seen anything remotely close to that emerge from them. On average, the average output in terms of deliverables is just, as I said, super bland.

Conor Doherty: If I may follow, but also if I could just say, because I mean, to be fair, that is your casual observation.

Joannes Vermorel: I don’t have statistics on thousands of companies.

Conor Doherty: You’re also a consultant, Eric. Feel free to reference your own experience as a consultant as well, please.

Eric Wilson: First, I mean, I would agree the average is bland. I would agree there. You know, 50% of the organizations, I said almost probably two-thirds in some cases, are bland. You do have the exceptions. It may not rise to the level of the paper that’s going to live for 50 years as the ideal strategy for most organizations, but you do get discussions as far as M&A, as far as acquisitions. There was a large CPG company that made the decision to pretty much let go of over 100 of their planners to be able to almost go lights-out type planning because they integrated AI and were able to facilitate the demand aspect. They let go of over 100 planners out of the decisions that were made in the S&OP process.

So there are those tough decisions that are made, but most of the decisions, I said even the average ones for the more mature, aren’t rising to the level you’re discussing. It’s more the level of acquisitions, new markets, those types of strategies that are really being defined. Brand, where are we going to be? What kind of player are we going to be? How are we going to position ourselves in the market? Those types of strategies are what we see discussed.

Joannes Vermorel: I agree with those being discussed. I agree with that. But also, I see that in fact, the decision has been made beforehand with a smaller number of people. And where my counter-proposition is that I see, and that’s where I see an anti-pattern, is that the meeting is a way to dilute the responsibility of the decision. So you see, there were like two or three people who were really championing something, but big bureaucracy, big company, big that. So they have this meeting. A lot of people in the meeting are poorly informed, not really following, checking their phone during the meeting and whatnot. And then the decision is acted together because, in fact, there are three people that are really pushing for it, and the rest basically don’t care.

And now the problem I see is that due to the fact that this meeting took place and that the decision was kind of acknowledged in the meeting, we have this big dilution of responsibility where instead of having one name to back this decision, we have 10 or possibly 20. And that’s why I say I’m very not convinced that those meetings, which is one of the main characteristic features of S&OP as it is practiced by most companies, you know, it’s a scheduled series of large meetings, that this does really contribute to making better decisions. Because I’ve seen when there is a strategic decision, it is made beforehand, and then what comes out of the meeting is just a dilution of responsibility. So it’s the same decision, but not made better, but made worse by just now having this fog of opacity on who is truly accountable for that.

Conor Doherty: To be fair, Eric, just to point out that in the same section “Efficient Decision-Making” of your book, you do identify defining decision ownership as a characteristic of efficient decision-making. So perhaps you could elaborate on that with reference to the challenge that Joannes has just mentioned?

Eric Wilson: Yeah, because I mean, the better you define roles and responsibilities, the better I understand what I’m bringing to the committees and meetings. The better I understand what I’m getting out of it, the better I can define roles and responsibilities. The more efficient the meetings and decision-making process is. So that’s what I was referring to.

From my experience, there are generally individuals driving the biggest part of the decision. Is it diluting? I’m not sure. What I have seen is consensus actually creates more accountability. When you have individual plans and decisions that people are driving, you end up with less accountability because if something goes wrong, you end up with finger-pointing. “They didn’t tell me that.” That’s generally what I’ve seen in most organizations.

Consensus, that one-number attitude that comes out of facilitated reviews and committees, actually builds more accountability. People are now on board. You have the drivers, but now we all have a consensus. What you end up with is more accountability as opposed to less. That’s typically what I see.

The other thing that those reviews facilitate is better communication. Companies have issues with communication, and these reviews help facilitate that. Not only do you have the driver and the approval, but you also have the consulted and the informed part of the meeting. That’s what makes up the committee. You have the driver responsible for bringing and driving the decision. You have the approval, which sometimes is not the same person who needs to approve the decision. You have the consulted, who needs to have their voice heard before finalizing the decision or strategy. And you have the informed, who understand what is being done and how it impacts them. That really makes up the committee.

Conor Doherty: Well, again, thank you, Eric. That touches on a point I wanted to transition into, which is more about the specific claims. Joannes, you previously said, and some of these are quotes and some are paraphrases, that S&OP is inefficient by design as it relies on information flowing through people. Essentially, communication, as Eric was just describing. You’ve referred to this as an obsolete way to transmit information in a modern company. So, two questions: one, can you expand on what you mean by that? And when you do so, please clarify what classes of information you are referring to.

Joannes Vermorel: The information I’m referring to is ultimately all the transactional information. First, let’s say there is no greater source of information about the company than its transactional history. If you look at information according to information theory, which is relatively modern, it can be quantified. You can quantify the amount of information you have, just like you can quantify how many kilograms you have of something. In terms of mass of information, the transactional information represents 99.9% of the information you have about a company. That is massive.

I’m not talking about a bakery that operates with one owner and three products. I’m talking about a billion-dollar-plus company with a spanning supply chain, many sites, many products, and many people. We are talking about something that operates at scale. The information is really there. Whatever opinion people may have and whatever extra insight they have about the state of the company, it is just a small fraction of this information.

The software we have nowadays doesn’t have the capacity to make really elaborate general intelligence analysis. What we have is the possibility to write special-purpose numerical recipes to solve narrow problems, such as controlling inventory replenishments. Having a completely different viewpoint on how I should even think about my sales data is still beyond what we can do with software. My take is if we want to think of this flow of information and organize it in a modern way, we need to acknowledge that the information is in digital form for the quasi-totality of it. There is no point in having people involved with the fine print of that. It’s just going to flow.

People are going to touch this information indirectly through layers of numerical recipes. Some numerical recipes generate decisions, some generate reports, and some generate all sorts of instruments needed for humans. There is a mediation. It is technically possible to look at the receipt line by line, but it’s pointless at scale. A modern view that I advocate, which is more software-centric, would embrace the fact that this flow of information is going to flow through software instruments of our own making. These instruments do not fall from the sky; they are created because people have thought about what sort of software instruments are needed, why, and how they are made.

If we meet together, you should not even be allowed to say, “I wasn’t aware that this product’s sales are spiking.” This information is available to everybody. If you don’t know, you’re not paying attention. It is not the role of another department to take your hand so that you look at the data available to everybody. Everybody has access to pretty much all the data of the company. That was the point of Bezos’ memo in 2002, where he said, “API all over the place.” Bezos was saying all the data of Amazon should be available to all the other people at Amazon, period. Your only responsibility when you’re part of a division is to make sure there is no data you keep to yourself to magnify your political power inside the company. Bezos said anybody playing this game would be fired within two weeks.

Now you have a landscape where the data is available. Obviously, there are plenty of problems. It’s not necessarily clean data, it’s not complete data, but making it accessible is just the start. If people have to meet, it is to discuss not any kind of decision-making but to improve their general understanding of the company. The meeting is not burdened with the idea that there should be any kind of deliverable.

In my experience, when you reach committee level, it is best if this thing is kept with a limited number of hours and the format is very free-form. Information should flow as unimpeded as possible. When people are away from the meeting, they can take their own responsibility and decide what they were going to decide anyway. This time, they don’t have the excuse of the committee to delay, postpone, or dilute the responsibility.

Conor Doherty: Okay, well, Eric, I do want to give you an opportunity to respond. There was a lot there, so I want to identify a few key points for response. A yes or no is sufficient. It sounded like you made the claim that a more software-centric approach would account for basically 99.9% of information flow within a company. Is that already the case?

Joannes Vermorel: Yes, it is already the case de facto. In no meetings do people start to just… I’ve read some historical documents about the emergence of General Electric. At the end of the 19th century, people had no choice but to read aloud in a meeting how much was sold, doing this and that. It would take hours and hours just to review. At the time, they had something like 150 products, and it would take them days to go manually because there was no other option. It was the pre-digital age. Nowadays, a company like General Electric is not having 150 products; they probably have 150,000 products, if not half a million. We are way past the stage where this information can even go through a human mind. It is incomprehensibly large. It has already been the case, whether people acknowledge it or not.

Conor Doherty: So that’s a yes. Thank you. And the second point was, S&OP essentially presents an opportunity to magnify political authority within an organization. An element of it presents people with opportunities to signal their position.

Joannes Vermorel: Yes, you know, that’s something that will just foster more bureaucracy. Bureaucracy just creates more of its own; it’s very difficult to contain. That’s also another of Parkinson’s laws: bureaucracies always grow. Salesforce will only grow if people are selling more; factories will only grow if they need to produce more. But bureaucracies inside a company will grow no matter what, even if the business is shrinking.

Eric Wilson: That part I agree with. I agree with that statement. Now, to his first point, he has me for about 90% of what he was talking about there. I’m actually going to give a story to help support what he said. The other part, it’s interesting what you were saying as far as the purpose of meetings and committees. I’ll take that and consider it going forward. I see other uses for it, so we’ll talk a little bit about that as well. First, the first part of what you said, I agree with 90% of it and I’ll give you a story to support it. This is a story that is in my book, so if anybody wants to get my book, you can read it for yourself. It’s in the Vanguard section, talking about where S&OP is going and how AI, data, and information can be utilized to enable a better S&OP process.

It came from a woman named Kim, and she gave this story. In a boardroom executive S&OP, they were using some type of generative AI, using internal and some external data. They put it into their systems in executive S&OP. The executive asked a question, “What should we do?” There were a lot of executives around the room. They put it into the system, and the system gave out a response saying, “Here are your options. Here’s what I would do. Here’s why I would do this.” The executive looked around the room to the other people. They just kind of looked at each other, and he said, “Okay, we’re going to do that.” We interviewed that CEO and asked him, “Do you think that is the best solution?” He was like, “I don’t know.”

So then the follow-up question was, “Why did you go with that?” He said, “Well, number one, it spoke my language. It was able to speak the language that I understood as an executive and operational, and it gave me a solution that made sense.” He said, “Secondly, it had my attention. Generally, what I would do in these meetings was I would ask these questions, and we have a lot of smart people making a lot of money around this table. They would go out of the room, do things, come back hours or days later, and say, ‘Here’s what we should do.’ By then, I’m already on two or three problems later. It had my attention.”

That is really the key. I think what you’re talking about when it comes to synthesizing information, processing the vast amount of information that’s available to make decisions in almost real-time, to meet the speed of business, we need to be able to enable systems and technology to do that for us, even in the boardrooms. I’m seeing a definite need for that, and we have use cases for that. I highlight one of them in the book as well, so I definitely see what you’re talking about. Now, the other point you’re making is if we go to that, then it’s just a report out or an information flow, or being able to level set everybody on where we’re going so they can make their individual decisions. That’s an interesting concept, and I will consider that. But what I’ve seen as far as mature organizations, they generally have three aspects to a mature meeting that’s an efficient type of meeting.

Number one is, “Here’s what everybody’s decided, the decisions that we’re pushing,” and it’s kind of a head nod to that of, “Yeah, consensus, everyone’s agreed, that’s where we’re going to go.” So that’s exactly what you’re talking about, whether it’s system-driven or people-driven. That’s the first part.

The second part is, there isn’t always agreement. There are some things that need to rise, either the systems or people cannot make the decision. We do need a committee to really look at that aspect. So there is exception-based decision-making. That’s a great part; you’re not making every decision as an exception-based decision. I see that maintaining even in the world that we’re going to, that we both kind of agree that it’s migrating to.

The third piece is then being able to articulate any changes that need to be made to the strategy going forward. It could be something as simple as, “We need to be more cash-focused because we have these covenants, and we see this downturn in the economy. So I know the strategies were this, but we kind of need to be more cash-focused in our strategies going forward.” So consider that with the decisions that are going to either be fed into people now or systems later.

So the three aspects are: here’s our plan, here’s the exception-based decisions that we need to make, and here are the changes we need to make. Now, whether those decisions are driven by systems and technology or by people, that I will concede on. But I think the other two parts are still a people aspect.

Conor Doherty: Thank you. If I can follow up, and I will come to you, Joannes, in a moment. I wrote down the three points you made. Again, I paraphrase, but one: consensus; two: conflict resolution; and three: articulating changes to strategies. That’s an interesting point and something I was going to come to later, but that’s a nice transition to something that Joannes had said. But I’ll ask you first, Eric. So previously, Joannes had made a claim that S&OP plans are ineffective given how slowly they are refreshed or revised. For example, people might only meet once a month or for some strategies once a quarter. This makes the processes ineffective in fast-paced business environments where conditions can change at a moment’s notice.

So my question back to you, and then Joannes, I’ll ask you the same question. With what you just described, I understand steps one and two, let’s just concede them for discussion. But step three, if you’re articulating changes to strategies once a month or like 10 to 12 times a year, how effective is that when things can change, like a container ship flips over in the Suez Canal? We need that decision today. We need to know what we’re doing now. We can’t wait another three and a half weeks. Eric’s on vacation, it’s a holiday for us. What do we do now?

Eric Wilson: I think it goes back to what he was talking about. Even if we have a system or we have capabilities of creating, “Here’s the information, here’s what we should do,” then individuals make decisions. It’s no different when you’re looking at a strategy. It shouldn’t be changing every day with the nuances of a port strike or something happening in another region that’s impacting you. Those are individual decisions that either people or systems should be able to make. Those are either operational or management-type decisions.

Where S&OP operates most efficiently is in the planning hierarchy, where you have the extended planning horizon where decisions can be made, not just trade-offs, where you’re making business decisions at extended planning horizons. Those shouldn’t change daily. Those are really driven by longer-term strategies of our organization. We have a cash strategy, we have a customer intimacy strategy. What are the strategies of our organization that really drive those extended planning horizon-type decisions?

During COVID, we saw exactly that there was a lot of chaos, a lot of uncertainty. What happened? Over 60% of the companies that we researched went to some type of weekly war room mentality with executives in these war rooms because things were changing so quickly, and they were trying to keep ahead of it. But what they discovered was that was inefficient. To your point, you couldn’t change that quickly. So most of those organizations went back to the monthly cycles with S&OP. Now, a lot of them maintained a weekly cycle for S&OE or master planning. If we want to talk about automating that, I’m all for that. That’s the management hierarchy, those are the decisions that should be made when you have the new information that is presented. Those are the decisions you shouldn’t do by committee. I would agree with that.

But a lot of those companies maintained an S&OE or master planning, but most of those companies went back to a monthly cycle because that’s where they found the efficiencies. We still have a strategy, we still need to maintain what our strategy is, and then execute those or manage those strategies in essence of what we do on a week-to-week basis.

Conor Doherty: Well, thank you, Eric. And Joannes, same question to you.

Joannes Vermorel: If I have to bounce back on what was presented, I think there is a subtle difference in perspective. I don’t even approach the decisions the same way. Just to give an example of how it’s done traditionally and how Lokad does it: we were discussing the cost of money. In a traditional meeting, there would be a discussion about whether we should reduce our working capital or allow it to expand.

Lokad takes a radically different approach. We just say there is a cost of money, an economic variable, and in the decisions that Lokad generates for inventory replenishment or production orders, this is one of the many economic drivers that is implicitly coded. This means that whenever these variables change, all the decisions generated automatically reflect the new reality of this economic variable.

If we decide that Finance is going to be in charge of maintaining this value, they can change the cost of money every single day, and all the decisions generated on the same day reflect that. They don’t need to talk to production or purchasing to get the effect of that. The synchronization is mediated through the software. There will be only one discussion about the fact that we have to agree that the cost of money is an economic driver that is company-wide and needs to be acknowledged by all. Once you have that, there are classes of discussions that are not needed. If Finance decides to bump the cost of money up or down, it is on them alone. Other departments will follow the rule set by Finance. If Finance sets a high-interest rate, then naturally, inventory will be sacrificed as a consequence of that economic assumption. They can revise it as frequently as they want.

The vast majority of things that are discussed can be isolated once as an economic driver of some kind. Then every department controls those strategic variables independently. The software removes the need for people to meet and discuss these things because synchronization and coordination happen through the software layer. It’s not exactly real-time, but it is super low latency, like minutes, and that’s just fine. Most committees are stuck with the perception that many things have to flow through humans, while if done more carefully, you can remove the need to discuss these things in the first place.

Eric Wilson: When you’re looking at decisions like the cost of money and holding inventory, you consider the cost implications. If we are heading into a recession and interest rates are going to increase, we need to decide what to do with the cost of money. On the flip side, we might want to gain market share during this time because we think there’s pent-up demand. We can create inventory and ensure customer intimacy to increase our market share. When we come out of the recession, we are positioned well in the market. Alternatively, we might decide to protect our cash and focus on margins. These are either-or decisions that often require a person or committee to drive them. The system can handle the math and efficiency, but the strategy piece still needs human input.

Joannes Vermorel: I disagree. Let’s stick to the example you’ve given of market share. The way we do it, we say $1 that you get from a new client is worth more than $1 from an old client. That’s it.

Conor Doherty: Please unpack that. I know what you mean, but others will not.

Joannes Vermorel: If you introduce an economic variable that says $1 from a new client is worth more than just a dollar from an existing client, what you’re effectively steering your decision-making processes into is whatever you can do to make those new clients emerge. And when they do, they will have extra dollars attached to them because that’s a way to modelize the fact that you’re investing to get extra market share.

And that’s fine. But you see, the way you can super concisely wrap everything into just how much is worth $1, what is the extra bonus that $1 of a new client brings for which period, and that’s it. And you see, that’s the thing where people say you have to either-or. I say no. When you have economic drivers, you can have a driver for the cost of money, one driver for the value of new clients, and you have those drivers. Different departments can steer those drivers in different ways independently. They don’t even need to talk to each other, and the resulting decision that you have is just the balance of all those drivers, costs, and rewards.

And again, you just remove, including for a push, to actually expand your market share that does not require people talking to each other on every single detail. You see, that’s the thing. That’s exactly the sort of things that I say people spend their time discussing things they should not even be discussing if they have the right setup with those economic drivers. Then it becomes people can take ownership of saying exactly how much is worth. And when you say that $1 of revenue of a new client is worth more than $1, it is not virtual. It means that marketing will actually spend this money somehow or through other functions, and it means that you will have an extra cost that will be incurred because you will have generated, let’s say, $50 million of extra revenues. But in order to do that, you will have generated $20 million of extra costs.

So, you see, the interesting thing is that you only need to have the economic drivers. The fine print of those extra $20 million invested to gain those clients is going to be ventilated across potentially millions of decisions, and you don’t really care. You see, you have this understanding of the structure of how the mechanic will play out. You don’t need to go into every single line to see exactly what is happening. You see a little bit where the difference is.

And I believe that’s really my frustration. Most of those discussions, I’ve attended those committees, they’re discussing things that with a proper setup, when I say proper setup, I mean modern software setup to drive those sorts of decisions, it’s mostly they’re talking about things that they would not even need categorically to discuss. That would be taken care of. So, it’s a big waste of time.

Conor Doherty: Eric, there’s a lot to cover there. Please, your response.

Eric Wilson: I think it goes back to an earlier part of our discussion we had that I really didn’t touch on then because I talked about maturity and journey, and you said that you’re opposed to anything that’s incremental. But that is truly what we see in an S&OP process. My whole book is really built out as a journey that organizations come through. I’ve went through the journey with organizations. I’ve countless companies I’ve worked with that have seen through this journey. It truly is an incremental process, and we lay it out and we’ve measured companies as far as where they are in the maturity. We can quantify where they are in that journey then as well, and we can see measurable results as they go through the journey. So I truly think there is an incrementality of this process. There is a journey of this process.

To that point then, I said 60% of the companies out there, and you see a lot of other research that shows the same exact thing. You know, 60% of the companies out there, I agree, they are doing things that are inefficient or you know, you can do otherwise in the meetings that they are doing, but they don’t have the enablers. They don’t have other processes, so they’re using this as an enabler to those decisions that do need to be made. So as they go through the journey, they’re either filling a gap or they are now going through a journey, and this is part of that incremental improvement to get to a place where they can be more strategic, and they are looking at things more efficient and the right things in the meetings in the future.

Conor Doherty: Well, thank you, Eric. And to follow up on that point and again just push this along a little bit, you talk some of the words you’ve used there, that the S&OP process is in fact a journey, it’s a maturity process, that there is a degree of incremental growth. Let’s just say that for a moment. Again, we’ve covered that, so let’s not go backwards. But going forwards, I am curious to talk about or to tie together what you just said about incrementalism and your discussion about cost and talk about something you said before, which is the potential opportunity cost of S&OP. So previously, as you talked about obviously supply chain decision automation, how that is basically the only way for supply chains to survive and for companies to thrive, and S&OP is an opportunity cost in that regard. So in what way, assuming that you still think that, in what way or please delineate the opportunity costs of the incremental S&OP maturity process that Eric is discussing.

Joannes Vermorel: The problem with this is that the top-level management of a company can’t be chasing an endless number of projects, of initiatives, of stuff. So they have to basically focus. I mean, out of necessity, they can’t really chase more than maybe a few dozen priorities, and that’s it. You know, there was this old joke about Microsoft which was focused on everything. But the reality is that it’s very, very difficult for a management team to push maybe more than two dozen initiatives. Now, which means that S&OP is going to consume one of those slots. That’s my point.

And what I say is that despite the fact that what I propose, the alternative to S&OP that I propose, is a massive reduction in terms of white-collar workforce because the alternative that I propose just mechanize the hell out of it. The mechanization, the savings in terms of workforce is just the tip of the iceberg. The real benefit is to have a superior version of supply chain for your company and also a superior version of what strategy means in a digital-first sort of environment. That’s what I’m saying.

And here I see those meetings as a little bit of the antithesis of that because they are not, you see, those things like S&OP have been around for decades. They will not be 10x better a decade from now. They are fundamentally limited by what people can achieve. I think the sort of practices are mature or they are known. They are not revolutionary. There is no, and just contradict me if you think I am misrepresenting what is presented, but I would say nowadays my take, after having read several books about S&OP, is that there is a general consensus with various flavors on how it’s done. It is not radically different. It is more a matter of preferences.

And whatever benefits we can expect, they are unfortunately relatively capped. The best S&OP is not going to produce the next SpaceX. It’s not going to produce the next Amazon. Those things will have very, I would say, capped sort of benefits. And I said the opportunity cost by going into an alternative paradigm that is very much software-driven is that that is the thing that can give you the 10x. That is something where you can do with two people what otherwise takes a thousand people and then have a reactivity that is like you react the next day instead of reacting the next month.

And I know that, for example, S&OPs just being able to react from one week to the next has been on the agenda of all companies for as far as I know at least two decades. They have never gotten that, and so they are still stuck. Many companies are stuck with actually quarterly plans, and they can’t even go to the monthly one, and they have been stuck for again one, two decades at that. So you see, I see the alternatives which are credible because it’s possible to make them work just give you things that we already know S&OP will never get there in terms of reactiveness, in terms of capacity to deal with unlimited amounts of information, the ability to just take advantage of all the nice goodies from computer science, from LLMs and whatnot. You see, that’s my point.

And the point is that this opportunity cost is massive, and by pursuing this classic S&OP, you’re just preventing yourself from making true breakthroughs into the alternative options. That’s the essence of the argument when I said the opportunity costs are dwarfing the savings that you can have by just reducing the manpower.

Conor Doherty: So Eric, just to summarize that for everyone, the direct and indirect costs of the process. The direct costs are what it actually costs in terms of salary, and the indirect costs are time, effort, and what you could have done with everything else. Your response?

Eric Wilson: So looking at the cost, I mean that we can quantify. We have quantified. Matter of fact, it’s in the book. I actually talk about what are the real costs of implementing an S&OP, what are the measurable benefits of S&OP, the ROI. It’s actually part of the book that I have. So we can quantify, we’ve done research on this in IBF. We looked at about 23 different attributes as far as the structure, cadence, who’s involved, the level of aggregation, the time horizon, the planning horizon. We look at different kinds, what you talked about, people call it different things, but what are the standard elements of an S&OP process. We measure those, the consistencies of, and we also look at how are companies performing as they go through the journey.

And what we found is we can now bucket people in four broad categories: emerging, just starting out; essential, those are the first half; next level, above 50%; and then vanguard, the top 15%. And what we found is that next level and above companies will outperform bottom line 2 to 3% improvement is what they see through the S&OP journey. That is real money for organizations. 1 to 2% top-line growth. They see a reduction in the uncertainty, whether it be supply or other types of uncertainty, be able to reduce those by 15 points. We see margin improvement. We see cash improvement of inventory. So across the board, they see measurable results. 2 to 3% to the bottom line for next level and above. There is real cost, and full disclosure, most emerging, it’s costing them money to do it. Most essentials, they’re getting minimum if any benefit when you’re at that essential level, which is half the companies out there.

So I fully agree, half the companies out there are not getting benefit from an S&OP process. But what we see, companies that go through the journey, there’s typically about an 18-month to 24-month ROI on the initial investment if you go through the process where they start reaching those next levels and they start seeing the benefits from it. You do actually see a return on your investment, and I have countless use cases to be able to actually show that.

Conor Doherty: So just to clarify, I don’t want to misrepresent that. Did you say 50% of companies do not currently benefit from their S&OP processes?

Eric Wilson: 15% definitely do not, and a large majority of the other 35% don’t as well. A lot of companies in the emerging and essential categories are breaking even at best.

Conor Doherty: Joannes, does that align with your sentiment?

Joannes Vermorel: My sentiment, again, is that I’m going to be much more pessimistic because, again, the problem is that you see when you tell me about ROI, it’s… let’s go back to this old analogy of, you know, horses versus cars. Yes, you can improve your horse by caring for it, by having a better horse saddle, etc. That’s the problem with those incremental improvements. They’re real, but the opportunity cost is so large. You see, that’s… people who are doing, “Look, we have improved the saddle for the horse, and we get a 5% improvement on what we can carry on the horse,” or “the horse is less tired, less prone to injury,” and whatnot. This is real, but you have an automotive on the side, and this is the thing that is like game-changing.

You see, people have a hard time understanding, but just imagine, as a quick anecdote, I already mentioned on this channel, but during the lockdowns, one of our clients, who had over a billion euros in inventory, a large European company, they had 900 users for them, and all of those people went on vacation paid by the state because, you know, European states were very generous, paying people to stay at home and not work for 14 months. So we had people that were staying at home for 14 months with, they were prohibited to work because otherwise the state would not provide subsidies for those people staying at home. So if you were staying at home, you could not work remotely; you had to be really on vacation.

And then, so we ended up in a situation with automation. We ended up with a team of like three people on the side of Lokad, five people on the side of the client, instead of the 900, to manage a billion-euro-plus inventory with all the manufacturing flow on top of that, still working for 14 months. You see, I see that as it is. I mean, the order of magnitude of change is just enormous, and the idea that you should be chasing a few extra percent where you’re not even sure that you’re breaking even, it’s… you see, for me, it’s not the right level of ambition.

Similarly, we had another client that, due to the lockdown, literally, it was a fashion company, they literally went under the equivalent for the international audience of Chapter 11. So they went, in 2020, under Chapter 11, and then when they got out, we had robotized several of their functions, and they were down to 90% of their workforce. And you see, that really changed the way you operate. You can do more.

Again, if you want to have an example of what a 90% workforce reduction looks like, look at Twitter in the year that followed. You know, those 90% reduction, Twitter introduced videos, they introduced half a dozen features that had been missing for over a decade because they reduced their workforce of software engineers by 90%. And there is really a causality. You see, I don’t think that people realize how toxic it is to have too many people on some problems.

And I see that, on the contrary, my deep belief is that the question is, can you have one person that can do as much as 10% prior? And if you can do that, the result, I have no counter observation in my experience. If you can achieve that, you will get things that are incredibly good out of that. Not just good in saving that you could do with the 10 people, but you will also have something that is of a better quality overall. It’s a very strange thing. Suddenly, you have more freedom, more reactivity, less bureaucracy. It is simpler, etc., etc. So again, that is a little bit my take on that. And yes, you can have improvements, but you’re passing on this kind of massive opportunity cost, which is, what about reimagining your company with 10 times fewer white-collar workers?

Eric Wilson: When you’re looking at the opportunity cost, I mentioned earlier a large CPG company located in the States that went through a process and decided to eliminate over a hundred of their demand planners. They didn’t need them; they could drive that by system. They understood going into that decision that their precision would most likely be lower, but the cost of all those people, net-net, would be a benefit to the organization. They are now hiring most of those back. It was millions of dollars they ended up losing because of that precision, not having the people. Not to say they didn’t implement some things correctly; technology is not there yet to do lights-out planning for that type of organization or most organizations yet. So we’re not there yet. Right now, less than 3% of all AI projects actually get implemented. There are countless use cases, but companies aren’t implementing them yet. That’s the reality we deal with right now.

The other reality we have to deal with right now is IBF research showing that 42% of companies use Excel as their primary operating system planning system. They’re not going to make the leap right now to where we need to go. I know you don’t like the word incremental, but you can’t just say, “Here’s this AI, we’re going to start doing this.” They have to transition to it. Right now, S&OP is a billion-dollar operation. I’m working with Microsoft right now; they’re developing their S&OP processes. Large organizations like P&G have established and are investing more and more every year. We see it used to be 15% of companies adopting S&OP; now that number is over 20% post-COVID.

So more companies are now adopting more S&OP processes. It’s a growing market, so not only is it a billion-dollar market, but it’s also growing. The realities we’re dealing with right now are that it is a process, whether it’s filling a gap or providing benefits of 2 to 3% to the bottom line. Those are real savings and real value for mature organizations. It is viable right now, and within the next few years, we’re not going to be able to eliminate it.

Conor Doherty: Thank you, Eric, because you’ve presented a perfect segue that I have ruined by acknowledging. It’s the question I’ve been waiting to get to, which is, Joannes, you’ve made the claim historically that companies can eliminate S&OP without negative consequences and should simply focus on improving data flow through software. To quote you, “I have met some large companies that decided to just terminate their S&OP division entirely; nothing bad happened.” So now, Eric has just mentioned Microsoft. They have some dollars; they’re pretty big. If they were to eliminate the S&OP process tonight, you’re telling me nothing bad would happen?

Joannes Vermorel: Microsoft is such a weird company in many ways. It has a profitability that is almost unprecedented in history. It is probably the only company outside of hard luxury that has more than 50% net margin. It is a very strange company. They produce excellent software, but they’re still running SAP, which is a competitor of theirs for their own ERP. It is bizarre.

So, I would not take Microsoft as a reference because it’s a software company. 90% of their revenue comes from digital assets that they are selling, not physical assets. They still have a pretty big supply chain with the Xbox and all the Microsoft hardware, so even if it’s small due to Microsoft being so large, the assets are probably a 10 billion plus business every year. So it is still a very large company. But to the underlying point, I’m very skeptical that it would actually hinder anything they’re doing. I do not see a lot of the headway of Microsoft, even as far as hardware is concerned, coming from very nice innovation, very nicely packaged products, usually superior software technologies, and better than them.

That’s, for example, the Surface that was created overnight. A billion-dollar business that was something like five years ago when they created Surface, the big ones for corporate meetings. So, you see, the way I say that is that it is not a game where those meetings are really creating the value. I do not see very superior execution for that. This is more like they have a very extensive range, they have decent enough software technologies. I know that people from the Linux camp would be horrified and say that everything that Microsoft produces is pure crap, but I would say it’s good enough. The equivalent of the stuff on Linux usually just does not exist, or when it exists, it’s not that good. Again, the equivalent of Excel on Linux is just scrap.

Conor Doherty: But again, just to press you a little bit on the overall point about any other company of a billion dollars or more…

Joannes Vermorel: Again, I did see that although the conditions were relatively extreme, it was usually where Chapter 11 or quasi-Chapter 11 forced people to do that. But again, I have seen companies hiring back people after firing a lot of people. As you mentioned in your anecdote, they were hiring planners and people dealing with daily decisions for inventory and whatnot. That I have seen as well. I’m not saying that those companies can just eliminate the people who are actually managing inventory or production planning. I’m discussing eliminating the people who are only focusing on S&OP, those whose function is to support those meetings, support company-wide forecasts, and that are only doing that. If you suppress someone who is managing the production schedule of the factory daily, you end up with a big problem if you don’t have a complete replacement production grade in place.

Eric Wilson: Even the companies that don’t have a structured S&OP process end up having something. Decisions are going to be made, so they may not call it S&OP. They may decide to call it something else. We have lots of acronyms out there that people like, but there are meetings happening even within those organizations. You don’t eliminate the process; you eliminate the structured S&OP process. I’ll guarantee you’re replacing it with something else. Now, you’re advocating replacing it with decision support and data flow and information flow. If they can do that, wonderful. But back to reality, 42% are in Excel. Most organizations don’t have the maturity. They’re replacing a structured process that’s proven and has benefits with something else.

Conor Doherty: Thank you. Just to clarify, Joannes, as Eric suggested, were you advocating replacing S&OP with something else? I wasn’t entirely sure.

Joannes Vermorel: I agree with the fact that there will be something else, but my point is that structure is not naturally good. It is not because you have a process in place that you’re going to get a good outcome. Sometimes having no process and letting things emerge gets you the best results. What I see is that you enforce a lot of rigidity, and it is not the sort of rigidity that is conducive to getting the best results.

There are plenty of industries, for example, the music industry, where you do not get a fantastically successful song by having a very careful process from the original idea to the finished product. It’s usually extremely messy. Those people live very weird lives, and it tends to produce good products. I believe this is rationalism at play when you want to add structure so that it doesn’t look dumb, to give a veneer of sense by putting things in neat boxes so that it looks under control.

My take is that remove S&OP, let those managers manage their time intelligently. If they don’t manage their time intelligently, just fire the people who are doing nothing. Those people will figure out what really serves them well and what doesn’t. You don’t need to have this uniform process across the organization enforcing those meetings. I’m not a big believer in those infrastructures that guarantee superior results.

Eric Wilson: You’re generally pessimistic. You’re optimistic that people on their own can figure out how to work and do things. That seems a little optimistic to me. What we found is there is a correlation between structure and the results you get. It creates more efficiency and better decisions with more structure. Now, I agree there’s a difference between structure and rigidity. You don’t want to be so rigid that you don’t have flexibility. You need flexibility, but you can still have structure and flexibility. I think those two can work concurrently.

Joannes Vermorel: Yeah, I mean, just to give you again an example that comes from a younger brother who has been working for a very large Danish company. I’m not going to give the name of the company for years, but people can look it up on LinkedIn if they want.

Just what S&OP looked like in process because that was, again, 40,000 employees, a large industrial company. He was part of sales. Every single year, there would be the S&OP process. So everybody would give their sales forecast. Every entry salesperson would give their forecast. That was like a super tedious exercise that would take two weeks for every single person of the sales teams.

And then obviously, because they want to exceed expectations, they would lower their targets. So they would undershoot massively. So then the boss aggregates. So first layer aggregates, and then they see that it’s undershoot, so incredibly low that they will have to retwist the number much higher, but not too high because otherwise it’s just ridiculous. They’re not going to lose 30% on their market share on every single market that they serve next year.

And so they would just rebump the thing up, but still, they want to exceed expectations, so they will still rester it down. And then it will go through the layers, and people will repeat this process of aggregating. It’s too low, bump it up, but not too high. And you end up with some super weird numbers. And at the end of this process that was taking months, and that was repeated every quarter, then you would have those sales projections that were complete nonsense.

And then people on production were doing the same thing, but they were doing it the other way. They would inflate their numbers because if they project more production, they would get more money to have more production capacity. So that was the opposite process, complete nonsense, also completely gamed. And then at the very top, they had an endless series of committees that would put those numbers one against another. And then at great length, they would finally agree on something.

And then people would be back. Sales would be in agreement of ignoring completely what was agreed upon because it was just complete nonsense, and production would do the same thing. And you see, that was the sort of thing where that is what happens with mature S&OP. And they had a mature S&OP that got to its maturity point at least 10 years ago. Maybe you would say it is done wrong, like the very wrong version of that. But my take is that that’s the problem with bureaucracy, is that bureaucracy tends to decay. So even if it was something that was done right at the beginning, the general understanding of bureaucracies is that they decay, and they decay fast. It’s extremely difficult to prevent a bureaucracy from decaying. So that’s, you know, that’s my take on that.

Eric Wilson: I think you watched one of my podcasts because I did an entire podcast about exactly what you talked about, especially during the budget cycle. People going into this, it costs the company so much money. There’s so much waste through exactly what you talked about because it’s utilizing people’s resources and time with zero benefit. They’re going in not trying to come up with an actual strategy result; they’re coming in playing games and going through the cycles of this game plan to come up with what’s going to benefit them. There is absolutely a waste of time and effort without even trying to get value out of it. I fully agree, and I’ve seen it. I’ve seen that exact case that you’re talking about in the S&OP process, but even outside of the S&OP process, I’ve seen the same thing play out. So I agree, what you described is inefficient and is done wrong. Yes, that is not the right way to do things.

But what you do see in organizations that have more mature S&OP processes, I mentioned Dr. Pepper, I mentioned Coca-Cola, some of those organizations, they’re not going through that budget cycle or that annual cycle that you’re talking about. It is actually integrated into the monthly cycle now, and it’s all about collaboration, having the right people work together, consensus driving through that one-number attitude, at least people are aligned. The other key is transparency. There is a transparent process, and that’s what really makes that mature organization. When you start trying to game it, when you’re trying to do things to benefit you, then it breaks down quickly.

Conor Doherty: I’m just going to push on a little bit from this point because I believe the producer is letting me know we’ve been going for about 90 minutes now. But it’s fun, it is absolutely fun, but I’m sure we all have things to do as well. So, again, listening to you, Eric, I’m really glad that you came, and thank you for coming. One, you’re nice to talk to, but secondly, you are quite open-minded and have alluded many times during the conversation to the future of S&OP. In fact, you touched on it earlier. I think you said the inevitable migration, that might have been the words you used, to automation, mechanization, and robotization. You said that Joannes was very optimistic about what people can do, and I’m curious, how do you hold both of these thoughts in your head simultaneously? That S&OP is great, we need to use that, and are you also optimistic about the future of S&OP, bearing in mind your appreciation of the general trajectory of technology and mechanization?

Eric Wilson: Yeah, I’m still optimistic. Things will fundamentally change, absolutely. I don’t deny that. I’m going to be in Europe next week talking at a keynote, and that’s going to be the essence of what I’m talking about. Things will fundamentally change from where they are, so we can’t continue the current process and trajectory we’re on and think that’s going to maintain us going forward. As technology enables more, as we get more capabilities, we will need to fundamentally change and enable new processes and new aspects. Those absolutely will emerge. What those will look like, I’m not fully sure.

What I project is we’re going to go more from even the strategies, more functional-specific strategies where we are right now, to more of the business side of things, more of the business efficiency as opposed to functional efficiencies. I think we’re going to go more there because we have the capability of synthesizing and being able to see how things more interconnect going forward. So I think we’re going to have more of those aspects.

The other thing that I’ve talked about is when one thing becomes commoditized, something else becomes a premium. Coding data is going to become more and more commoditized. What’s going to become the premium then is the communication, the objectivity that is outside of systems that we need to then start appreciating. Being able to look at things and question even what the outputs are saying, and how to question and what to question and when to question, those are going to be the skill sets going forward that are going to be incorporated into some type of business efficiency planning or the S&OP of the future. Those are going to be really what we’re going to be designed to. But absolutely, I think we’re migrating. We definitely see change happening, and I’m not going to be ignorant to that. It’s not going to happen over the next few years, but we will see it migrate, and we will see a fundamental difference in what we have in the future.

Conor Doherty: Thank you. And Joannes, before you respond, I just want to pick up on a certain point that was mentioned there. Eric, you talked about how naturally new processes will emerge in time, that’s the general trend. But something you said earlier, Joannes, was essentially there’s a limit on the innovation of human-to-human communication or how well information can flow through people. So bearing in mind you said that, what is your response to what Eric has just said about the future of S&OP?

Joannes Vermorel: My take is that when you go through an incremental perspective, you always add. You see it as it is fundamentally an additive process. And where I see the true disruption and true, you know, leap forward is when you subtract. You know, it’s a very abstract sort of way to think about that, but if you want to have a beautiful, beautiful example, look at the three rocket engines, first generation, second, and third designed by SpaceX.

I mean, I invite the audience to have a look from generation one to generation three, where they stand. The Generation 3 is way simpler. It’s literally an order of magnitude less pipes, tubes. The first thing looks like a mad scientist experiment. It has hundreds of tubes. The third generation has something like half a dozen visible tubes. It is… the one looks like… it has this sort of Dyson vacuum cleaner vibe, you know, very elegant, simple, everything has been removed. The first one was just like mad science experiments with like a steampunk vibe with so much complication.

So you see, my point back to this evolution, I say it would be… I don’t know where we are heading. I really don’t. I just say that the easy thing that you can think of is just adding, but the true, you know, leap through the things where you progress massively is when you subtract. So my question is that would be think about long and hard about what will be removed and how from S&OP 2.0, even if it’s, you know, even if my take is incorrect. Maybe the take would be it would be S&OP 2.0 not with more but with less because it is done smarter, differently, something else.

Conor Doherty: At this point, I have nothing further to add. Is there any point that you wish to pose to each other unmediated by me? Is there anything burning on your mind you want to ask, Joannes?

Eric Wilson: Well, as far as Joannes, I think we’ve ended on agreement. Neither one of us fully knows. It may be a simpler version. I’m all on board with that, but it will be a different version going forward. I think we find consensus on that. We built consensus. So we have consensus on that. When you’re looking at the future of demand planning, you said 90% can be eliminated. Is your view we can actually eliminate a lot of the people and be able to automate more of the business going forward? And if that’s the case, how do we deal with the equilibrium of wage versus consumer? Because we still need people to buy things, and businesses don’t buy things, but businesses aren’t consumers. How do we balance that going forward?

Joannes Vermorel: So, here the take is that I’m very much in agreement with the Schumpeter perspective on economy, which is creative destruction. You know, 200 years ago here in Paris, the job number one was, and it was a quarter of the population, it was people who were carrying water. That was a quarter of the population. Obviously, running water has eliminated these jobs, and people, those… that part of the population just didn’t die out of starvation because they had no more jobs.

You know, whenever there is something that just mechanizes, then those people are liberated into the job market, and the job market finds its way. You know, there is no limit in how much… I have a daughter at school. They have classrooms of 30, sometimes 35 children. You know, I would be very pleased if they had classrooms of 15 children instead. There is… when you go to the hospital, you have to wait sometimes in Paris for hours because they are short on people. You know, there is no shortage of having more nurses, more school teachers, more everything.

Companies that operate supply chains where nowadays the white collars exceed the number of blue collars, it’s very strange. You know, mechanization has been working so well that factories are almost empty. And now I have many clients where a factory can be run with 20 people, but you need 100 people doing Excel spreadsheets on the side of the factory to keep the factory running. It’s very strange.

If we go back a century ago, people would have been completely baffled. They would think, “Oh, we need three engineers for 100 workers,” and nowadays that’s the opposite. You have like literally hordes of people doing spreadsheets. I completely agree, spreadsheets do govern supply chains. In contrast, the blue-collar workforce has been extensively mechanized.

So we have all those people who are doing back-office jobs. My take is that, yes, for those, I would say, back-office clerical jobs, you know, I’m very specific about the sort of job: back-office clerical. So it doesn’t see people who do not see a client, people who do not see a supplier, people who do not physically do something for the company. Those will most likely, that’s my take, 20 years ahead will be reduced by up to 90% with software technologies. That would mean millions of people.

The market will find its way. Do not think of these sort of dystopian scenarios where those people are just left to starve, having no resource. They would just find a way. Just because, again, that’s the beauty of the general economy, the free market, is that there is no limit on how much stuff people want. So if you have free resources, the market will find a way to just allocate those people to the most paying jobs that will be just different things.

I think that the interesting thing, really concluding, is that it is the first time this is happening to white collars. Digitalization started in the late 70s, never eliminated white collars. We have more white collars than ever. But now I think we live in a society of a bubble of white collars where people think it’s normal if 90% of the population sits behind a desk instead of doing something physical. I believe it is very strange. It is something that is very much like the first section of the 21st-century sort of view. Historically, it has never been like that.

My take is that it could flow back to a different sort of society where we go back to something that is much more balanced, where blue-collars, and I say that in a non-pejorative way, you know, I know that in the US you can have plumbers that do up to $150,000 a year. Those people are not poor; they are making a very good living. It will just balance back to something where there will be a lot less of those clerical back-office bureaucratic sort of jobs. That’s just my take for the next 20 years.

Conor Doherty: Does that satisfy your curiosity?

Eric Wilson: Semi. I look at it a little bit differently because I’m not sure this, as far as automating white-collar, we’re not mechanizing. We are innovating new forms of decision-making and AI. I think it’s going to be radically different going forward. That has the potential to be different than other types of innovations that we’ve seen historically.

The mechanization of certain tasks and processes that people do, we may need to rethink the way we have workforces. We can get philosophical, but we may look at minimum income type of scenarios because businesses are becoming more efficient, they’re making more money, and they have a tax base. There are a lot of ways we can look at it, but I’m not sure that we’re going to have the society we have in the future that we had in the past.

Joannes Vermorel: Countries have been changing a lot, and I think the US is no exception. The US has evolved tremendously over the last two decades.

Conor Doherty: in 24 hours.

Eric Wilson: Did you see our election a couple of days ago?

Conor Doherty: All made the same joke at the same time.

Joannes Vermorel: I think expecting change is normal. Since the very start of the Industrial Revolution, there has been constant change. 300 years ago, it was like 3,000 years ago. That’s the interesting thing.

Up to 300 years ago, you could go back in the past 3,000 years, it would be mostly the same thing. Then with the Industrial Revolution, we ended up in a spiral of increasing knowledge that just improved everything. So yes, I think we can all expect change. I think the worst that we can expect is pure stagnation. It’s probably the worst that a country can expect.

Conor Doherty: So then, Joannes, to provide some sort of storytelling symmetry, if you were to collapse all of this down into a memo, which I know you love, what would the memo for this long podcast with Eric be? And then Eric, I’ll ask you as well.

Joannes Vermorel: If we want to summarize and be fair to the position of Eric, it would be: there are methods where you can achieve your company-wide synchronization, company-wide consensus, that are less wasteful than probably what you’re doing. That I would agree. My counterpoint, and that’s a claim, and I’m not going to go into the detail of supporting this claim, is that while you can get those benefits, there are even bigger benefits and opportunities by challenging the idea that you operate through people by going through a software-mediated sort of paradigm. But then I will leave this claim unsupported because otherwise, I’m going to take yet another hour to go into that.

Conor Doherty: Thank you. And Eric, your Jerry Maguire-style memo?

Eric Wilson: My Jerry Maguire reality of where we are now is S&OP is increasing in adoption. It is filling a void right now of what we have. You do see benefits as you incrementally improve the process. If you want to understand what it is and get on that journey and incrementally improve, buy my book.

Conor Doherty: Joannes, thank you for your time. And Eric, thank you very much for joining us for so long as well. It’s been great having you.

Joannes Vermorel: Thank you very much, a pleasure.

Conor Doherty: And thank you all for watching. We’ll see you next time.