00:00:00 Introduction to IBP and its purpose in business
00:02:10 The promise of IBP: unified company vision and strategy
00:04:25 Challenges of coordination in large, geographically distributed companies
00:06:10 The opportunity cost of focusing solely on IBP plans
00:08:10 IBP’s limitations in adapting to market changes
00:10:15 Critique of IBP’s narrow, time-series based approach
00:13:00 Consensus theater: misaligned incentives in IBP meetings
00:16:00 Bureaucratic inefficiencies and hidden costs in IBP
00:18:15 The role of software in amplifying IBP’s inefficiencies
00:21:45 The problem with using proxies in IBP
00:24:30 Forecasting accuracy as a proxy metric issue
00:27:15 Challenges of low-resolution decision-making in IBP
00:30:00 The high cost of low-resolution decision-making in supply chains
00:33:00 The importance of granular data in supply chain decisions
00:36:00 The limitations of IBP’s software-driven, process-oriented approach
00:39:30 The need for human decision-making outside of IBP’s rigid structure
00:42:00 Using technology to manage granular supply chain data
00:45:00 The problem of over-complicating IBP processes with software
00:47:30 Economic perspective vs. IBP’s bureaucratic pitfalls
00:51:00 Final thoughts: reducing bureaucracy and focusing on real value
Summary
IBP promises “one company, one plan,” but often delivers bureaucratic comfort instead of better decisions. A single time-series plan narrows vision, hiding market shifts that don’t fit the grid. Cross-functional alignment becomes “consensus theater”: predictable tug-of-war, expensive meetings, then teams ignore the plan anyway. Monthly cadence adds decision lag, while aggregation creates low-resolution averages that conceal real problems. IBP software amplifies process compliance and proxy metrics, not outcomes. Better governance means aligning on durable economic drivers, then letting machines handle high-frequency, granular trade-offs.
Extended Summary
Integrated Business Planning (IBP) sells executives an appealing promise: one company, one vision, one plan—revised monthly—so everyone marches in the same direction. That promise made historical sense when large organizations struggled to coordinate across distance and function. But the modern version of IBP often confuses “having governance” with “making better decisions,” and then charges handsomely for the confusion.
The first hidden cost is opportunity cost. A “single operating plan” typically means time-series tables aggregated to what fits on slides and in meetings—volumes, dollars, categories, regions. But markets change in ways that don’t respect your grids. When the future is forced into a narrow template, management becomes blind to transformations that matter: not just how much demand, but what kind of demand, where it shows up, and how fulfillment expectations evolve.
The second cost is what might be called consensus theater. Different functions have different incentives: sales benefits from sandbagging forecasts; manufacturing benefits from optimistic volumes to justify capacity. The result is predictable tug-of-war—expensive in time, thin in insight. After months of argument, many teams proceed as they would have anyway, sometimes discarding the “agreed” plan entirely. The company pays for choreography, not clarity.
Then there is the information problem. Monthly or quarterly cadences build delay into decisions, as if decision latency were not part of lead time. Worse, a meeting cannot transmit much information—humans can only process so much. You end up making high-stakes choices with low-resolution tools: averages that hide collapses and surges inside categories, and segmentation schemes so arbitrary that different groups would classify the same reality in incompatible ways.
Finally, software-driven IBP magnifies bureaucracy. Large organizations naturally drift toward process compliance because it is safer for careers than owning outcomes. Add workflow software and you industrialize that drift, multiplying proxy metrics—like forecast accuracy or “requests answered in one day”—that can be optimized while real performance stagnates.
The alternative argued here is not “no communication,” but different communication: align on relatively stable economic drivers and definitions of service, then let machines refresh high-frequency decisions. Companies don’t win by unanimous plans; they win by making better bets, faster, grounded in economic consequences rather than ritualized paperwork.
Full Transcript
Joannes Vermorel: I’m good.
Conor Doherty: Good. I’m good.
This is Supply Chain Breakdown, and today we’ll be breaking down the hidden costs of Integrated Business Planning, or IBP for short. I’m Conor Doherty, Communications Director here at Lokad, and to my left, as always, Lokad founder Joannes Vermorel.
Now, before we get started talking about IBP, let us know down below: one, are you a fan of it? That doesn’t mean that you love everything about it, but you know—you’re an advocate, a devotee perhaps. And secondly, do you think IBP has a place in an increasingly AI-driven supply chain?
Alexey is our producer today. Get your comments in, and Joannes and I will discuss those a little bit later. Now, let’s get started.
Joannes, before we start, a definition to set the table. So today, when we say IBP, what we mean is the original—I believe I’m using the Oliver Wight definition that you cited in your recent article.
So we mean, quote: “The formal way a business is managed with a monthly exception-driven process and a single operating plan that links portfolio, demand, supply, and finance over a 24 to 36-month horizon.”
Now, there are many flavors of IBP the same way there’s many flavors of S&OP. Perhaps you have slightly different interpretations. That’s what we’re working with. Alexey, please drop that definition in the chat.
If you want to challenge that, that’s fine, but for today, that’s what we’re working with. And with that, Joannes: IBP—one plan approach, a single united vision. Before we get negative, what does that promise to executives when they undergo an IBP process?
Joannes Vermorel: The gift of it is: one company, one vision, one strategy, one plan—and a completely aligned execution. That’s this “one-one-one-one.” Let’s everyone together in the same direction to achieve that.
And indeed, if you go back to the emergence of the mega-corporation in the 19th century, yes, it proved to be very much a challenge. As soon as you reach a company that gets geographically distributed—so you’re sufficiently large that everybody doesn’t even fit in one region—it becomes very complicated.
You can have people who produce stuff, other people who are actually selling other stuff, and you have mismatch. Some people are very good at selling something, but the people who are producing are not catching up, etc. Those misalignments can create a lot of complications.
So it was this idea that they needed—besides the idea of having this unity—they needed to have a glue. And this original approach of IBP was to structure the thing that would give cohesion to the company.
That’s how you had all those processes with monthly planning iterations. There are specific ways to bring everybody on board with a certain cadence. And yes, the fundamental idea is to get everyone on board, on the same page.
So literally: one company, one vision, one strategy, one plan. The plan will be revised on a monthly basis, and then convene everyone, and use this plan as a commitment so that every single piece of execution is aligned toward this plan.
Conor Doherty: Okay. Well, this actually came about from a blog that you wrote last week, and I just want to quote you from that blog. You said, and I quote: “I do not dispute the value of governance in relation to IBP. I argue its sufficiency.”
So, in a way, those words kind of make you sound sympathetic, or even appreciative of IBP in some looks. Where do you see it actually providing help—or at least the appearance, or patina, of help?
Joannes Vermorel: The fact that the company, at some point, needs to think about having a shared vision for the business obviously makes sense. Do they need to have a shared strategy, a little bit more detail, etc.? Yes, probably.
Should people talk to each other occasionally? Probably. Should manufacturing never talk to sales? Probably not. So there are plenty of things that are just very reasonable.
It is reasonable that you try to have channels of communication, and some of them are a little bit established so that you can have that. It makes sense. That’s why I say I do not dispute the value of governance, because if you say brutally “no” to all those ideas, that would mean: what, you want to enforce the silos?
Obviously, this is not what I recommend. My criticism is that fundamentally this can get ritualized, and it can over time devolve—this IBP process, and it usually does—into something that is very inefficient.
And in fact, the truly useful channels of communication end up outside of those sorts of rituals.
Conor Doherty: Well, on the note of inefficiencies, that’s what actually led to the idea of the hidden cost of IBP. So, as a summary for anyone who’s just joining—because the numbers are going up—view from 30,000 feet: what is the hidden cost of this single operating plan approach?
Joannes Vermorel: Maybe one I didn’t even discuss in the blog post is the cost of opportunity. Oh, huge cost, yeah.
The problem is that when people say, “We need to have a plan,” because that’s going to be the deliverable of IBP, the problem with this idea of one plan is that suddenly that makes you blind to all the stuff that does not really fit in your planning view.
Because when people say “plans,” they mean something extremely specific. They mean a list of time series, aggregated at a certain level, and that’s going to be the plan.
So that’s going to be the volume of activity per week and product category. It’s going to be the turnover per week per activity, or per region. You have your hierarchical organization of your company, and that’s going to have a certain level of disaggregation, but not too much.
It’s going to be very aggregated overall. We are going to take raw volume—volume in units, or in dollars. But the problem is: what if the market is changing in ways that are slightly off your grid?
Business can transform among many dimensions. For example, we have been working with many clients, and one of them observed a transformation where they’re a B2B distributor.
It turned out that increasingly—they had first a relatively rapid move in the last decade toward e-commerce, where their professional customers went to order online instead of ordering in the stores. And then what initially—they were ordering to have their deliveries made on site, where they have the construction works.
But gradually they realized: “Oh, it creates so much complication, because sometimes the construction site doesn’t even have enough space to store all the things that are received.” So increasingly what they wanted was to have their big orders being delivered and ready to be taken, but stored nearby.
And that was a subtle, but a very important transformation. That was essentially something like click-and-collect.
This is something where, again, if you think in terms of the classical time-series perspective, it’s not the sort of thing that you will see.
And when I say “problem of opportunity,” it’s that it forces you to look at the future through the narrow lenses of the planning instruments that IBP makes available.
That’s why I say we really have to differentiate: getting together and discussing the future, and thinking long and hard about the future, versus establishing a consensus on the time series. Those are very, very different propositions.
When I say the opportunity cost is that it locks everybody into looking at the curves, and then you have a tug-of-war to steer the curve upward or downward, instead of thinking: is there a deep transformation in our business that really changes the very structure of the market?
And suddenly we should not even be thinking anymore the way we were, because it requires a deeper transformation.
Conor Doherty: Well, the thing is—so you’re… I like the way you’re framing that. Because again, we take an economics approach. You’re talking about opportunity cost.
Another dimension to the economics perspective is trade-offs. And on that note, there were many comments on your post—and again privately, when I spoke to some people about, “Hey, we’re going to do this talk, is there stuff you want me to ask?”
A general consensus was around the idea of trade-offs. I’ve amalgamated—Chat, here are quotes that people said to me; summarize this in a general sentiment. It was: “IBP is not perfect, I know that, obviously. However, I have to bring together many cross-functional teams, and IBP at least provides some direction. It’s a common language, not a perfect tool.”
That’s the language of trade-off. So are you sympathetic to that problem, or is it orthogonal?
Joannes Vermorel: Yes, you need to have trade-offs. But again, do you come from a world in the 19th century, where the only instrument that is available to you is essentially pen and paper and discussion?
Or do you live in the 21st century, where you can use a computer to compute those economic trade-offs?
My argument is that, when it comes… You see, we had the opportunity, and my criticism on opportunity was the sort of numerical framework you have: those time series, they lock you into something that is very narrow-minded. You have very strict lenses on what you’re looking at.
And here you have another problem: despite the fact that you put yourself in a place that is numerically very constrained, you don’t even leverage that to actually do an economic calculation of those trade-offs.
That’s also something that I found very lacking. It’s very incomplete in this regard, where your plan is not explicitly financial, it’s not explicitly economic. It is just about projected volumes. That’s it.
You don’t factor into it any kind of explicitly—again, those are not factored—all the costs and benefits that are going into those numbers. They are at best buried under the assumptions that let you come up with those macro plans.
Conor Doherty: Well again, I just want to point out: I like the use of the term “incomplete.” Again, it’s not completely trashing the entire concept; it’s “we can do better.”
Now, on that: just for the sake of today, when we were talking about hidden costs, there were four. Now you mentioned opportunity cost. But the three that you outlined in the blog—one of them was “consensus theater.”
And you were arguing—that’s a lovely phrase, by the way. I love the way that you write that, “consensus theater.” But in IBP context, and for practitioners listening: what is that, and how is it a hidden cost?
Joannes Vermorel: The consensus theater is that, again, the incentives of the various people—the various teams—are not aligned with the long-term interest of the company.
If you’re part of a sales team, what you want is to sandbag everything. So you want to lower all the projected numbers so that you can exceed expectations.
If you’re manufacturing, that’s the opposite. You want the numbers to go high in terms of projected volume so that you can justify the investment into capacity, so that when you will observe the demand, you have the capacity to serve.
So you have those tug-of-war that are taking place. That’s why I say “consensus theater,” because it is a show. It’s very time consuming, and in the end every single division still does a little bit the way they want.
The resource allocation ultimately—the plan is big numbers that will consume a lot of time for everybody. Then everybody keeps doing pretty much more or less what they were already doing before.
That’s the crazy thing. People will fight for months over a plan, and then sales keep doing pretty much what it was doing, and manufacturing keeps doing what they were doing, etc.
I’ve seen so many companies where it takes like a quarter to come up with the IBP forecast. Then when I ask the sales team, “What do you do with that?” “Oh, we just throw it away.”
When I ask people from manufacturing, “What do you do with that?” “We throw it away.” And what about the people from warehousing? “Oh, we also throw it away.”
So you see—again, the problem, I go back to my first point about opportunity cost: those discussions are sterile. People have this tug-of-war where you are driving this one forecast up or down. It does not really inform anything or anyone.
As a result, that’s why I say it’s theater: it’s mostly futile, performative.
And if we have to go back to what is real, what is real typically are the allocation of resources. So what is this thing connected to the allocation of resources, and the fact that you want—if you have an economic perspective—maximize the rate of return on your allocations? The answer is: none.
It is extremely disconnected. That’s why it becomes also very bureaucratic, because you have those recurring meetings. It’s going to involve a lot of people.
Conor Doherty: So you’re outlining there two points, because they don’t necessarily all fall under opportunity cost. There’s the direct tangible costs of salary while people sit in a room and discuss things.
And then there is the opportunity cost of what you could have done with the same resources, had you allocated them differently.
Joannes Vermorel: Yes. Exactly. And again, that’s also when I was talking about the devolution of IBP—which I see pretty much everywhere—is that yes, you need to have communication channels between sales, manufacturing, warehousing, transportation, etc.
It turns out that the most important discussions will take place outside of those meetings. For me, that’s a sign that all the executives of the company will bypass that. It’s discussed outside of the meetings, and that’s where the discussion will be made.
The recurring meeting is still happening, but nothing of substance is decided there.
Conor Doherty: But then the natural follow-up, as I’m sure many people listening are thinking, is: if I don’t use IBP, and communication is necessary, then how do I execute those meetings? Like, what is the alternative, then?
If I throw away—or I massively amend—my IBP process, what do I do?
Joannes Vermorel: First, when people say IBP nowadays, it’s not just a process. It’s actually a software workflow that is sold to you by a software vendor. This is where the bureaucracy can get answered.
Companies have bureaucratic temptations. It’s already very hard in a large company to make sure that whatever initiative you have doesn’t devolve into some kind of bureaucratic undertaking. It’s extremely difficult.
The way large human organizations operate is that it is safer—not more profitable, but safer—for the people inside to make sure that everything becomes process-oriented as opposed to outcome-oriented.
If you belong to a large organization, you do not want, as a personal individual, to have anything in terms of outcome that gets attached to you, because it’s a big risk. Organizations are very risk averse.
So if anything is outcome-dependent, that means that when the outcome is bad, it’s going to be bad on you. You take a substantial amount of risk as a middle manager or upper manager in this corporation.
So the temptation—which is extremely strong, and that’s what is happening—is: instead of being outcome-oriented, you become process-oriented. And then you would just say: “You know what? We did it by the book. Literally. We followed the process, all the steps, and so we are 100% compliant,” which is super safe.
“Oh yes, the outcome was terrible, but we were perfectly compliant.”
That’s the temptation. And then if you add software—because this is what I was saying—this temptation to have everything devolve into process-oriented compliance, instead of outcome-oriented, is already super strong. But if you put software on top, software is an amplifier.
People don’t realize that software tends to amplify the good and the bad. With IBP: if you go back to the version in the 80s—which is very much a managerial thing, where software is very secondary—it had this problem of bureaucratic devolution.
If you put IBP software on top of it, then you are massively amplifying this problem. You speed up this devolution massively, and you end up with things that get ritualized even more. The workflow of the software steers the nonsense.
Conor Doherty: When you say companies can become outcome-disoriented, they should be oriented in the direction of making more money, but depending on what kind of process they undertake—for example IBP—they might become more process-oriented.
Is that suggestive of your argument against proxies? So the idea of: instead of “I’m trying to make more money,” I’m trying to improve my forecasting accuracy, and then that’s what my teams are oriented towards. But that might not necessarily move the needle financially.
Joannes Vermorel: Yes. Proxies are a corollary of the sort of process-orientation devolution.
If you are looking for the outcome, then you have the perfect metric, which is rate of return, or long-term amortized profit of the company. This is very straightforward.
Now, that’s not that common, though. I mean, it is for any company except the largest one, it is actually fairly common.
The problem is that if you are in a large company, you have so many things that you do not control. Again, the problem with outcome is that people will say, “It went bad, it’s your fault.”
So what you will want to do is: you want to take a metric that is under your control. That means that you will take a proxy. That’s going to be a metric that reflects a step in the process.
And again you would say, “Oh, it’s my outcome,” but it’s not an outcome. It is literally some kind of compliance inside the organization.
If you say we want to have forecasting accuracy, people will pretend that is something that is equivalent to an outcome—business performance, real dollars coming into the company. No, it’s not. It is absolutely not.
It is a numerical artifact that is completely disconnected from business performance.
Why are those artifacts introduced? Because bureaucracies love numerical artifacts.
Conor Doherty: When you say “numerical artifacts,” again: numbers that don’t necessarily express the underlying performance of a company.
Joannes Vermorel: Yes, exactly.
Conor Doherty: Okay. Just for anyone who doesn’t know: it’s a number that kind of reflects the bureaucratic perspective of this team or this team. Forecast accuracy, for example?
Joannes Vermorel: Forecast accuracy, or for example: percentage of customer requests addressed within one day.
Because if I tell you, “Oh, we address 99.99% of the customer requests within one day,” people would say, “Great,” and accept that we just deny the request, all of them.
Conor Doherty: Well, technically, yeah. So we are super fast at whenever there is a customer complaint—we are just going to reply instantly. We don’t care. And so we can tick the box that it was handled.
Joannes Vermorel: Exactly. But that’s okay. Obviously, that’s an extreme example of Goodhart’s law, which I know you talk about a lot.
Most companies—even if they wouldn’t admit it publicly—would agree: yeah, that is a problem. Essentially, if we make a metric and that’s the yardstick, people orient towards that because that’s their job performance.
And people may not even realize this because those things happen gradually. If we take, for example, forecasting accuracy, what will happen is that people will start looking at: “Oh, our accuracy is so bad at the SKU level.”
So we are going to aggregate that. First, maybe not the day level, but maybe the weekly level. Still bad. Let’s go for monthly. Monthly is still bad.
We should aggregate by month and by product category, or maybe by month and by region. Then the accuracy starts to look better.
But you see what I have done: I started from the real problem—which was: how much I need to allocate for every single SKU every single day, in terms of inventory, production capacity, etc.—to a forecast value and a forecast accuracy that is measured at a super aggregated level, both time-wise and scope-wise.
Yes, the metric looks better when you do that, but you’re further and further away from anything that would be like the outcome for the company.
That’s why I said you have this sort of devolution. For example, when you introduce with software those IBP software layers, they will give you so many more numerical artifacts.
They will multiply the amount of proxies that look scientific but are mostly nonsense. Another one would be “forecast value add”—layers on layers on layers of things that have no correlation whatsoever with profitability of the company.
But thanks to those layers of software, you can have all of that. Then you can declare some people responsible for those things, and you just inflate the bureaucratic structure.
Conor Doherty: Time-dependent—we may come back to the comments on FVA, because I think it is something to cover.
But to push on more directly—because again, I want to stay on track: hidden costs.
Another one that you expressed—and it’s one actually I quite like—it’s not too philosophical, because it’s actually a very practical one. It is, as you point out: literally the way you view supply chain.
So the idea of: if you have a standard IBP process, you’re looking at monthly meetings. What that means is: unless there are exceptions—and of course there are exceptions—you might have an ad hoc meeting.
But as a general principle, if you’re meeting once a month, you’re deciding things together once a month, which means you’re viewing your supply chain in 12 incremental steps.
You challenge that robustly: the idea of monthly versus daily, or possibly even hourly. Please outline that. How is that a cost?
Joannes Vermorel: A good supply chain decision is an informed decision. If you go for a monthly process, then you add one month of lag.
It’s as if people don’t realize: when you think of lead times, you need to think of all the delays, including the delay to come up with a correct decision.
It’s not just how much time it takes for your supplier to deliver, how much it takes for the factory to process and build. It is also how much time it takes to get to the purchase order, get to the production order, get to the inventory allocation order.
Here, IBP is extremely slow. One month of lag, and very frequently it’s every quarter.
It is very slow. And then I also challenge how much information can actually flow during those meetings.
If you start assessing in terms of information theory: how many bits of information—how many Shannons of information—can flow during those meetings? The answer is maybe a few dozen, and certainly not more than a hundred bits of information.
There’s a ceiling to the informational perspective: how much information can flow through those meetings.
By the way, that’s something I detailed in my recent book. There is a chapter on information. You can quantify the amount of information in a mathematical way.
Realistically speaking, if you do that about a meeting, the amount of information that can flow—we are talking of something that would be less than a kilobyte, much less than a kilobyte.
So not only it is very slow—it lags—but the resolution, information-wise, is extremely poor as well. That is a real practical limitation.
When you think in terms of flow of information of machines: not only you can have things that flow in literally seconds from one side of the company to another, so instead of having a latency of one month plus, you can have something much shorter.
On top of that, the resolution—the amount of information, the granularity of the information—can be, again, if you use machines as opposed to humans and meetings, several orders of magnitude higher.
When I say “several,” I mean like six orders of magnitude higher. It’s going to be literally a million times more informative.
On one side, you have not even a kilobyte of information that can flow through a two-hour meeting—and I’m being very optimistic when I say a kilobyte.
On the other hand, you can have literally tenths of gigabytes, if not terabytes, of data that can flow during the same amount of time through your systems.
Conor Doherty: Well, again—because I do want to ground it as much in terms of costs—when I bring up the idea there of economics, I know that, let’s say to take Lokad’s perspective, it would be:
For example, you manage a hundred million of inventory at any given moment. Your supply chain manages a hundred million dollars worth of goods.
Our perspective would be: every single one of those dollars, if you were to disaggregate it, has its own potential economic return.
So if you take that at a monthly cadence—once a month you come in and you decide in a meeting—let’s even say it’s a four-hour meeting, and let’s even say for the course of four hours you maintain glucose levels and you can focus and you’re superhuman, and your team is superhuman.
Realistically, how granular can your economic discussion be about that large of a portfolio, and then say with a straight face, “We take an economic perspective”?
And I’m not being critical—I want to be clear. I’m not being critical. I’m just saying: that’s unreasonable to ask of people.
Joannes Vermorel: Yes. And what happens in the end is that you’re very low resolution—time-wise and scope-wise.
That’s why you have this one month of lag—and one month is, again, being optimistic. Even if you have those monthly meetings, it doesn’t guarantee that something pressing is going to be absolutely treated in the next meeting.
It can lag further.
And then also in terms of resolution—how granular are you? You can have, for example, a category that looks stable, except that in this category you have some products that are collapsing demand-wise while some others are exploding demand-wise.
If you just look at the average, it looks fine—stable. In fact, it’s completely made-up stability, because you have things that will cease to have any appeal in the market soon, while others may completely exceed your capacity to serve customers.
Again, if you just aggregate everything, you don’t see those patterns anymore. You just see an average that looks like: “Oh, everything is fine.” It’s just buried in those averages.
Conor Doherty: You used the term “made up”—essentially like it’s a made-up kind of sense of control.
Maybe “fragile” might be better, because it breaks. Everything’s made up, obviously—we choose a perspective. But some will be more robust and some will be more fragile.
Joannes Vermorel: No. When I say “made up”—if I say I’m counting the number of units of this book, yes, this is a tangible reality. We can agree that there are two units on the table.
Then if I bring a second book, I would say it’s a different product. So if we count, that’s going to be different units.
It might be arbitrary in the sense that it’s a man-made object and that’s another man-made object, etc. But it’s not completely arbitrary.
Now, if I decide that on the shelf where I have all my books, I’m going to do a segment for supply chain and a segment for logistics—yes, that becomes… those two segments have quite an overlap.
That’s why I say it starts to be quite made up, because it’s really about your judgment.
That’s why I say the problem with those bureaucratic processes is that when you start looking at the way they segment things, it is incredibly arbitrary.
You will have segments that are very arbitrary. As a consequence, you can have a lot of instability where your segments look stable, but the reality is that you have tons of products that jump from one segment to another.
You have tons of clients that jump from one segment to another. You also have segments that may not represent anything truly real or meaningful about the market.
That’s why I say it’s very made up. When you start slicing and dicing according to very arbitrary rules…
A litmus test would be: if I take 100 people and I recount those books on this table, they will all end up with the same assessment. No question.
If I take my bookshelf and I ask people to classify those into four categories, and I take 100 people, they will come up with 100 classifications.
That’s why I say: is it really made up or not? Would different teams, if they were to redo the work, end up with the exact same thing, or would they end up with something completely different?
Conor Doherty: Well, this actually—and I will push on in a moment because there are some audience questions, both from LinkedIn and YouTube—we’re on both.
But this ties into a point that you’d made before about S&OP. I think there’s a bit of overlap in your feelings for both IBP and S&OP.
It sounds like the limitations you describe are what happens when you take a high-dimensional problem—one that really should have computer and automation intervention—and try to express it through the mind of humans.
Talented humans, very well-intentioned humans, but the human mind can only conceive of this problem at a certain level of complexity, and that is going to be lower than a computer.
So then you have these very made-up, simple, low-resolution perspectives. That’s more or less what you’re saying?
Joannes Vermorel: Yes. Exactly.
In the specific case of IBP, when people say IBP nowadays, they really mean something that is driven by some kind of enterprise software workflow.
That’s where I say it’s a net negative in this specific case. If we go back to the ambition—one business, one vision, one strategy, one plan—suddenly you reduce the amount of meaningful communication between the relevant parties.
It is not the same to have the salespeople talking face-to-face with an open discussion with manufacturing—no agenda, just brainstorming and making sure we’re on the same page—as opposed to having some kind of tug-of-war on numbers that are initially forecast but are going to be turned into commitments.
It’s very, very different.
The second one—the workflow operated by software—I think is incredibly restrictive. It’s time consuming, and in the end it does not really inform the various parties that much.
On the contrary, it simplifies; it removes all the substance and the gotchas.
That’s why I see very frequently companies reinvent on the side something that would be closer to the original idea of IBP, which was to have those lines of communication so that we stay on the same page.
But most of the important ideas that need to be conveyed are not going to be numerical.
Your software infrastructure is going to take care of making sure that everyone is on the same page as far as numbers are concerned. Unlike the 80s, now everyone can see the sales, and everyone can see stock levels, etc.
You don’t need to talk to the salespeople to know if sales are coming in—you see that; everybody can see that on the system.
Same thing: salespeople don’t need to be talking to manufacturing to see if they have extra inventory. The stock levels are in the system.
For those purely numerical communications, you don’t need to go through the teams—you can literally go through your software infrastructure.
Conor Doherty: All right. Well, my last question before going to the chat was going to be about “one plan” versus “priced stance,” what you propose. But actually the first question tees it up.
This first question is from Daniel, on YouTube: “As I understand the perspective you propose, it is to make the lens wide, to then make it narrow again and choose the most probable or lucrative option. If that’s the case, wouldn’t I come to one final plan anyway?”
Joannes Vermorel: No.
That would be true if we were to discuss something where the number of options was small.
For example, if you’re a venture capitalist firm and you want to make one investment in a startup a month—yes, you can. Then you will have one decision, and that will be your capital allowance for this month.
You can do your meetings and debate and come up with a conclusion. By the way, that’s exactly how venture capitalist firms operate: the partners get together, they look at a deal, and they say, “Do we go in or not, at which price?” And bam—they strike the decision.
Okay. But supply chain is not like venture capital.
You don’t make one capital investment per month. You make tens of thousands per day.
So the idea that you can discuss the rate of return of the resource allocations together—you will never get to that.
That’s why IBP devolved into something that is very low resolution, time-wise and scope-wise, precisely because you can’t get into the nitty-gritty detail.
So what I say is that instead, you should be focusing on getting a consensus on the economic forces—the economic drivers—the way you should even structure those things.
How do you, for example, think about quality of service for customers? We need to have an agreement on that.
It’s agreeing on the very perspective. Sales might have something to say; manufacturing might have something to say on that. So when it comes to agreeing on those economic perspectives—what does that mean for the company—this is where people need to bring together, discuss, and get to some kind of consensus.
But as a rule of thumb, you want to have consensus on the stuff that does not change—or at least does not change quickly.
For example: if you are a grocery store, the market has been… If you’re a brick-and-mortar grocery store, the quality of service for a hypermarket today is not fundamentally very different from the quality of service for a hypermarket in the 70s.
So there is stability.
If you’re an aviation company, MRO, and you want to avoid aircraft on ground—again, it has been like half a century. The problem of avoiding aircraft on ground is very much stabilized.
Yes, the Airbus A350 did not exist 50 years ago, but when you want to avoid those aircraft on ground, whether you’re dealing with an aging Boeing 747 or a brand new Airbus A350, it’s going to be pretty much the same.
So what I’m saying is that you want to take the time to focus on what doesn’t change, as opposed to have the humans involved in chasing ever-changing figures.
If you want to get to an agreement on something that the computer can refresh every single day, that is a little bit nuts. You want to have the humans out of the high-frequency computational loop of your systems.
People are never going to stay up to date with that.
Conor Doherty: All right. Well, thank you, Daniel. I hope that was edifying.
This next one is somewhat long, but I’ll chunk it. This is from Boris Yushmanov. Hi Boris.
“As for IBP, I’m fully aligned with Lokad. The definition can differ, but for me the real question is: how is IBP used? For example, running day-to-day operations through IBP—bad idea. Using it to take the long view, assess strategic scenarios, make big decisions, give shareholders a structured view quarter to quarter—that makes sense. As such, IBP works as a strategic compass, not as an operational approach. Your thoughts?”
Joannes Vermorel: We are getting back to the original 1980s vision for IBP.
The question is: is it what is being sold through the major competitors of Lokad who are selling IBP solutions? That’s the key question.
I agree with the feeling. Now, if we look at what is being sold under the IBP umbrella: is it what’s being sold, and will those solutions deliver that?
My take is: absolutely not. Absolutely not.
That’s the challenge. My humble take is: I’m not too sure if we go back to this original 1980s vision, refreshed for the present day—where the information is flowing through the application landscape of the company—I’m not sure if something like IBP really needs a software infrastructure and workflows.
I would rather suggest to have it as part of the culture, where people routinely talk to each other. They try to actively break the silos, but not necessarily too much codification.
Because if you codify those sort of things, you make it very boring, and then suddenly the information does not flow as fluently, as easily as it should.
I’ve seen in many companies: if you want the information to be digestible by humans, the information needs to be made for humans.
I’ve seen many situations—large corporations—where the documents that are being produced by this sort of bureaucratic entities are so incredibly boring that nobody can really get to the bottom of those documents unless you’re a McKinsey consultant paid lavishly just to do that and endure the boredom.
Those documents are so incredibly boring that first, people cannot really manage to get through them. Then even if they manage to get through them, they will immediately forget everything.
So that’s the problem of the effectiveness of communication.
My humble experience is: if you make things very repetitive, very codified, you make them incredibly boring, and in terms of communication, very quickly people tune out. It becomes paperwork.
People skim the thing, and it does the exact opposite of what people intended to do in the very beginning.
Conor Doherty: All right, thank you.
A DM just came through, so I’m not going to say who it’s from, but I will begin with the compliment. Big fan of the show—thank you, thank you. Good to hear that.
I’ll start again: big fan of the show, but I’d like to hear more about the politics. Our IBP has strong internal sponsorship. Realistically, how do you handle awkward discussions when you try to sell this economic view to a room of IBP owners?
That’s from the C-level, by the way. I’m not going to say who, but that’s from the C-level.
Joannes Vermorel: The thing is: first, the economic perspective is very tough.
But the reality is: you can bury your head in the sand, but you can’t avoid the consequences of your decisions—or lack of decisions.
For example, if quality of service is killing your business, and when you put a dollar on that you realize, “Oh, this is literally going to get us bankrupt within a decade”—yes, it is mildly terrifying. But what is the alternative?
Just pretend that you can badly treat your customers and everything will be fine?
Or if once you put those dollars you realize that there is a business unit that is really doing a bad job and it should, most likely for the greater good of the company, be terminated—because the other business units are just fine—should we just pretend everything is fine and wait until the entire company blows up due to the damage caused by this one business unit?
Yes, the economic perspective makes it very blunt, very tough.
But my experience is that on the contrary, it tends to diminish the respective powers of politics, because suddenly there are facts, and people have to face the facts.
Unless your company is completely dysfunctional, most managers and executives in a company operate on good faith. Occasionally you have bad faith actors, but most of the time people can be a little bit selfish, but still operate on good faith.
Thus, having those facts suddenly makes people face the hard decisions.
Now, the problem that I have is: if I am facing an IBP committee—by the way, at Lokad, when we are trying to sell Lokad, we are not going to try to sell them to those people, because ultimately, if we get our way, these sorts of things just disappear.
That’s the case in most of our clients: those sorts of bureaucratic approaches to planning, they just disappear. They’re not needed.
Not in the 21st century, where you have machines to do the super intense low-level calculations. You don’t need to have those planning meetings anymore.
That’s a little bit like: how do you sell an automotive to someone who is the general stable manager—the guy who is managing the stables for the horses?
If you want to sell an automotive and you’re going to pitch it to the committee that supervises the stables across all the locations of the company, I guess it’s going to be a very tough sale.
But really, here, in order to make sense, we have to think of the endgame: those allocations of resources, and we need to optimize them economically. That’s my message. With computers, it can be done well.
Conor Doherty: Thank you.
I’ll push on. This is a question from Alfonso. There are many parts of this, so broad comment:
“Isn’t the real problem with IBP that the framework is often misused? Everyone can see it isn’t working, but no one wants to say it. In that case: one, how should we build the business plan so it is owned, clearly designed, and aligned with strategy? And two, how do we turn the outcomes of IBP meetings into concrete daily actions that truly connect to all functions?”
You kind of touched on that just now, I think.
Joannes Vermorel: First, when people say it’s misused, I very much disagree with this sort of idea.
We are in an environment where IBP is nowadays pretty much driven by software workflow. So here we need to blame who is responsible, which is a software vendor.
For example: if after typing queries on Google for 20 minutes you cannot find the headquarters of a large company, who will you blame?
Do you think that is your Google searching skills that are lacking, or is it Google who is just giving you crappy results and does not understand that you’re looking for the address of this company?
I’m making the case up—Google is actually excellent at finding the geographical address of the headquarters of any company—but just to press the point:
A good software solution will steer the users toward success.
If the vendor can just say, “We are blameless, you guys are just idiots, you misuse it,” no. Sorry. It doesn’t fly. Not in enterprise software.
As an enterprise software vendor, you have to look at the “pit of success,” so that the default—if you just let yourself roll down the slope—you will roll down into success by just letting gravity do its work.
If the natural course of events is the pit of despair—where people have to do heroic efforts so that they don’t end up into the pit of despair—for me this is a very bad software solution.
Good software solution will effortlessly drive you to success. Bad one: you will have to be heroes not to end up with failure.
Back to that: when people say it’s badly used, I would say: not really. My take is that it’s poorly designed, and I’m talking about the software.
It creates a lot of bad things. It amplifies the bureaucratic nature of the undertaking. It makes it even more bureaucratic. It makes it even slower, etc.
Now, second point: how do you turn the outcome of IBP meetings into concrete daily actions?
The problem is: what should be the outcome? That’s where I really disagree on the idea that your outcome should be some kind of business plan.
I very much disagree. Or some kind of plan—I also very much disagree.
It’s more information. The problem is that it’s going to be way too low resolution.
Your business plan is going to be a joke. It’s going to be something with 10 numbers, and that’s it. Again, extremely low resolution.
If you are a venture capitalist and you have one investment to make per month, being super low resolution is acceptable, because ultimately you say yes or no for one deal, or maybe five deals that are on the table, and that’s it.
But here, if you say you have your plan or your business plan: what is going to be the granularity of this document? It’s going to be super coarse. It’s not going to be something very detailed.
Thus, you’re going to average everything like crazy.
Just think what it means to be low resolution. Imagine I say: “I want to install a new McDonald’s in Paris.” Do I want to say “Paris” as a location? It’s crazy.
In Paris, there are terrible locations. There are streets where you have no people going by. You have places that are insanely expensive, but there is no pedestrian traffic.
You need a resolution that is incredibly fine-grained. If I want to make a decision—where do I put my McDonald’s?—I need to decide exactly this street, this place, and at this price point for the rent.
That’s incredibly high resolution.
What you’re doing effectively with IBP is as if you were saying, “We are going to put a McDonald’s somewhere in Paris and be done with it.”
Or in inventory terms: send some stock to Paris, versus send this pen to that store at that price point, and this pen to that store at that price point, at this date.
That’s when you say high resolution: you’re talking about that level of granularity.
Conor Doherty: Like that store, at that price point, at this date. When you get—when you say high resolution—you’re talking about that level of granularity.
Joannes Vermorel: Correct. Yes. Exactly.
That’s why bureaucracies, who have been operating in this theater for so long, they tend to forget over time that what they are doing is incredibly low resolution.
They don’t even realize. It’s not even part of their perspective that it would be possible to get to the nitty-gritty detail of every single SKU, every single unit.
But that’s where all the games lie. Ultimately, money is made or lost at the lowest level.
McDonald’s makes money one burger at a time, not through a macro investment and saying, “We need to invest in this region.”
If you look at the capital allocation, they’re much, much more granular, and that’s where the profitability.
Conor Doherty: Exactly. Could not put it better myself.
So to push on—again, this one’s from Alfonso. To be fair, there’s a lot of back-and-forth between people actually in the chat, so these questions are sort of an amalgamation or a composite of thoughts.
But: if we keep areas like transport and supply chain strictly specialized and separate, how are major business changes—like a new product line, a new plant, etc.—supposed to be communicated and absorbed, if not through IBP meetings?
Joannes Vermorel: Strategy, really.
But again: should information flow through your company through meetings? I say it shouldn’t. It should not.
Eventually, high-level insights should be forged and refined through meetings, not fine-grained information.
Just to give you an example: we have clients that are introducing thousands of new products every month. How does the information flow? They don’t need meetings for that.
There will be teams that introduce the products. It’s going to be listed in the ERP, and then all the information will flow automatically.
A product can be introduced, and will be stocked, distributed, priced, promoted—etc. All of that is orchestrated largely through the application landscape.
So you don’t need to have people talking to each other all the time.
We have clients that have large retail networks. They open and close stores every week. They don’t ask Lokad. They just do it.
We see in the system that there is a new store listed. We push the inventory.
We see that there is a store that is listed for termination at this date. At this date, we stop pushing inventory.
Again, we don’t need a meeting. The information is flowing through the application landscape.
In a modern corporation that is digitalized—and all large companies nowadays are digitalized, they have been since the late 90s—99.9% of the information flows through their IT systems, not through word of mouth and people talking.
So for me, even if we go to this ideal of the 1980s, in this 21st century world, the point is not to convey basic facts about the company. All the facts are recorded in the business systems: your ERP, your MRP, your WMS.
All of that flows through the application landscape. You don’t need to have a meeting. Everyone has access to this information—or should.
Some companies have dysfunctional setups and the information is not as accessible as it should, but fundamentally this information should flow across the board without having any people involved.
If there is a meeting, it is to see: how should we even think? In which markets are we even operating?
Is the market changing in ways that redefine how we should be looking at ourselves?
In the history of markets: if you look at the history of Nokia, they completely missed the smartphone era. They were the leader on phones until they were no more, because they missed the smartphone.
Same thing for Blackberry, etc. It’s an entire series of manufacturers that completely missed the transition toward smartphones.
If we go back to things like IBP: this is the place where you want to discuss those things, where you need to make sense of things that are not going to be reflected in the facts.
Things that elude your codification of facts. Things that are beyond your application landscape.
If you go there to discuss mundane information, you’re just wasting your time. Your application landscape does that much more efficiently than people can do it.
Conor Doherty: Okay. Well, actually speaking about that, I’m just trying to amalgamate some thoughts here.
One of the questions is essentially more like a case study. It’s putting you on the spot, but: could you comment more or less on the kind of savings that the perspective you’re advocating right now—how much of the hidden cost could you remove?
I’m just going to take an example: an omnichannel retail situation—40,000 SKUs, 200… sites, etc.
Obviously it’s rough numbers, but when you’re talking about leaving money on the table, low resolution, lacks granularity: are we talking about thousands of dollars?
Joannes Vermorel: No.
In a situation like that, you would have maybe a team of 15 people who are managing the planning—planners.
First, we are talking about pretty much eliminating those 15 people. That’s just a saving in terms of wages.
Those people can be redeployed to things that are much more useful, such as managing the suppliers more closely. It’s the value add of those people.
So we eliminate the jobs—not the people. If they have experience, instead of babysitting numbers, they can do things that are much more valuable.
But the idea of having people who are essentially massaging numbers and doing meetings—those things are just gone.
In terms of cost: 15 people, the order of magnitude will be a million a year of pure cost that you can remove.
Plus on top of that, you have the solution by the vendor that unlocks the savings: probably, even for relatively smallish companies like this one—mid-range—that would be half a million a year of ongoing cost, maintenance, that would also be removed.
Then you would have the setup fee—probably like two million. It’s once, in theory, that should be able to last a decade, but in practice, in three years it’s going to be completely obsolete due to the pace of change in the software industry.
So we are talking—easily—half a million a year in this kind of mid-range for the IBP software solution that you don’t need; one million for the people who are kept busy as privileged users of this IBP solution.
Then on top of that, one million for people who are interacting with this IBP process: all the directors that have to be involved for meetings that are not very productive; a lot of people in finance, in marketing, etc., who need to touch base with this IBP process routinely.
Most of this time is completely wasted.
So I would say we are talking about maybe two million a year of pure friction cost that can be eliminated by just removing that.
Again, you’re not losing much, because fundamentally this thing is just not delivering.
For me, the proof of that is that during the lockdowns—2020, 2021—I saw many companies who actually took advantage in Europe of the incredible governmental subsidies that said: if your white-collars stay at home—no VPN, they are not allowed to work—if they stay at home without working, you get a subsidy to pay for those employees.
It was ludicrous, but a lot of countries in Europe, including France, did it.
What I saw is that companies—those people were deemed non-essential workers—and the companies operated for like 14 months without those people, and it went just fine.
Even without Lokad: you remove those non-essential white-collars for 14 months, and the company is doing just fine without them.
For me, that proved that it’s just bureaucratic overhead. Companies are under the impression that they need it, but in fact they don’t.
Do an Elon Musk thought experiment. Think for a minute: how would Elon Musk manage your company?
Elon Musk purchased Twitter—now X—and he fired 80% of the staff. Out of the 20% that remained, half quit. So they are like at 90% less staff, and the thing is just working just as fine as it did before.
They had so many layers of people who were just not doing much.
That’s a crazy thing that I see in supply chains nowadays: we have optimized incredibly everything that is blue-collar, but when it comes to white-collar, we have armies of clerks that do comparatively very little.
That would be my message: don’t be afraid to drastically reduce those layers. They are not what gives you your competitive edge.
Conor Doherty: All right.
Well, I’m out of questions, but I did want to finish somewhat symmetrically.
We have a quote from Mark Twain. I’m going to read it to you, and then you give me your closing thoughts:
“A company doesn’t win by agreeing on a plan. It wins by making better bets faster with eyes wide open to the future’s unruly shape.”
You wrote that. It was not Mark Twain. Just to be clear, that was you.
What do you mean by that, and how does that tie all of this together?
Joannes Vermorel: Ultimately, it is what you do—and that is tangible—that matters.
We have to go back to what is real. If you put something on the shelf and there is a client that comes and gives you dollars for those things, it is real.
If you have a fantastic designer who invents a new design and it outcompetes because it’s so much cooler than the competition, it is real.
If you decide to place an order from an overseas supplier, it is real, etc.
So again, the things that are real are very important and consequential.
My message—especially with things like IBP—is: beware. It is incredibly tempting in large organizations to get lost in busy work.
It is so tempting. Organizations are constantly learning how many ways there are to create busy work.
Every single technological buzzword ends up generating so much busy work—things that are not real—where people invent proxies, invent metrics, and processes, and silos, and whatnot.
First, you have bureaucracies that create silos, and then you invent a process to break the silos, but through this process you’ve recreated all the silos, etc. All of that is just not real.
So my message would be: really think it through, through what is real, and what are the economic consequences in dollars for the company.
If things feel very fuzzy, very meeting-ish, very bureaucratic, slow, you should have red flags.
Think: “Okay, this thing is just not creating value.”
Do thought experiments. If you were Elon Musk and you take over your own company and say, “You know what, I’m going to do the same thing with only 10% of the people,” which ones would you keep, and why?
That’s a thought experiment. It might not be very wise—maybe the whole company would just blow up—but it may give you a sense of what is truly essential versus what is essentially complacency.
Conor Doherty: All right. Well, Joannes, we’re out of questions. We’ve been standing here for 70 minutes, so we’re now officially out of time.
Thank you very much for joining. I always have.
And to the rest of you, thank you for attending, for your DMs, your questions, the lively debate in the comments. It’s great to see.
As always, if you want to continue the conversation, you can reach out to Joannes and me privately on LinkedIn. We’re always happy to talk—even about supply chain, whatever you want. Philosophy—we talk about it all.
And on that note, we’ll see you next week. Have a good week, and get back to work.
