Alignment and Decisions in Modern Supply Chains
Whenever I talk with supply chain executives, I’m struck by how many different “theories of everything” they are offered. Some emphasize process, some technology, some relationships, some culture. Among the serious contributors to this landscape, John Gattorna stands out. His work on dynamic alignment, customer behavior, and tailored supply chains has influenced a generation of practitioners, especially in Australia and Europe. His language is very different from mine, and so is his starting point. Precisely for that reason, putting our views side by side is a useful way to clarify what I actually believe supply chain is about.
I recently tried to set down my own perspective systematically in my book Introduction to Supply Chain, where I frame the field as the business of making profitable choices about physical goods when the future refuses to sit still. But a book-length treatment can become abstract. So in this essay I want to do something more concrete: place my view in dialogue with Gattorna’s, highlight where I agree, where I diverge, and how these positions might fit together in practice.
Two different starting points
Gattorna begins outside the firm. His “Dynamic Alignment” framework starts from the idea that the ultimate reference point for an enterprise should be the external marketplace: the customers and end users it serves. From there, he argues that sustained performance comes from aligning four elements: the characteristics of the market, the company’s strategy, its internal culture, and its leadership style. In his books, especially Living Supply Chains and Dynamic Supply Chain Alignment, he describes supply chains as “living” systems driven by the behaviors of people—customers, partners, and employees—rather than mere flows of boxes and bits.
He then proposes that most customers fall into a small number of dominant buying behaviors. To respond effectively, companies should not operate a single monolithic chain, but a small portfolio of distinct chain designs, each matched to a behavioral segment. At a high level, he distinguishes continuous replenishment chains for collaborative customers, lean chains for efficiency-driven customers, agile chains for demanding customers who value speed, and fully flexible chains for customers seeking innovative, highly tailored solutions. The emphasis is on fit: what kind of relationship and service pattern is appropriate for each segment, and how should culture, processes, systems, and leadership be arranged internally to support that pattern?
My own starting point is deliberately different. I begin inside the firm, at the level of concrete decisions: how much to buy, where to send it, which orders to accept, what service promises to make, at what price. I treat supply chain as a branch of applied economics focused on physical goods under uncertainty. The central question, for me, is: given our limited capital, capacity, and time, which combination of actions gives us the best risk‑adjusted return, once we account for the fact that demand, lead times, and operations will never behave exactly as expected?
Where Gattorna frames supply chain in terms of aligning behaviors and cultures, I frame it in terms of allocating scarce resources under uncertainty. His unit of analysis is the behavioral segment and the corresponding chain archetype; mine is the repeated, granular decision—millions of small bets made every year on stock, transport, and service promises.
We are looking at the same phenomenon from opposite directions.
Customers first, or cash first?
Gattorna’s work is unapologetically customer‑centric. He insists that you should begin by understanding how different customers buy, what they expect, and how they behave under stress. Companies, he argues, routinely over‑serve some customers and under‑serve others, mostly because they never did this behavioral segmentation properly. Once you see that different customers are playing different games with you, it makes little sense to offer all of them the same logistics experience, the same response times, the same collaboration rituals.
I have a lot of sympathy for that critique. In my own practice, I have repeatedly seen companies offer gold‑plated service to low‑margin accounts, while quietly starving profitable but “unimportant” customers because the organization never revisited an old segmentation. Where I differ is in what I treat as the final court of appeal.
For me, the ultimate scoreboard is cash—more precisely, the stream of cash flows the firm can reasonably expect to generate, adjusted for risk. This is not because I value customers less, but because I don’t know any other way to compare conflicting demands honestly. A demanding but strategically important customer may deserve premium treatment; another demanding customer with poor margins and erratic behavior may not. From the inside, their pleas sound remarkably similar. What allows you to tell them apart is their long‑run contribution to the firm’s economic health.
Put differently: Gattorna takes the customer as the “fail‑safe reference point”; I take the company’s risk‑adjusted economics as that reference, with customers, suppliers, and internal constraints all feeding into that calculus. In practice, of course, these views are not mutually exclusive. If you consistently misread what customers value, your cash flows will eventually testify against you. But the emphases matter. A framework that stops at “alignment” can mask the fact that different alignments have very different economic consequences.
How we think about uncertainty
Both of us see volatility as normal, not exceptional. Gattorna writes about an operating environment that rarely returns to equilibrium; he emphasizes that companies must learn to live with constant change in demand, competition, and regulation. His practical response is to design multiple supply chain configurations that each embody a different stance toward uncertainty. A lean chain economizes on cost because demand is regular and relationships are loose; an agile or fully flexible chain accepts higher cost in exchange for speed or bespoke solutions in markets that are volatile or innovation‑driven.
In other words, he addresses uncertainty largely at the level of architecture. For each behavioral context, choose a different combination of buffers, capacities, and collaboration patterns, and keep adjusting these as the market shifts.
I tackle uncertainty one level down, at the level of the numbers themselves. Rather than classifying a chain as “lean” or “agile” and deciding qualitatively how much slack to keep, I want to express the underlying randomness explicitly: the probability that demand for a specific SKU in a specific location next week will be 0, 1, 2, or 20 units; the probability that a container will arrive in 18, 25, or 40 days. This probabilistic view of demand and lead times then feeds into a decision model that prices stock‑outs, obsolescence, and working capital in money. This is the same probabilistic, economics-first stance I develop in the “Forecasts, plans, and the illusion of certainty” section of Supply Chain as Economic Bets in a Market-Driven World.
The result may look superficially similar—more inventory where demand is erratic or lead times are risky, less where things are stable—but the logic is different. Instead of defending a qualitative label (“this is an agile chain”), I am forced to be explicit about the tradeoffs: for this item, in this week, how much profit am I sacrificing if I protect myself against the 95th percentile of demand rather than the 80th? That kind of reasoning is tedious to do by hand, but relatively cheap to encode in software and run across millions of decisions every day.
You could say that Gattorna manages uncertainty through top‑down pattern recognition and configuration, while I try to manage it through bottom‑up probabilistic modeling and optimization. One looks at the chessboard and chooses an overall style of play; the other evaluates move by move, given the position.
Technology and the place of people
Nowhere is the contrast sharper than in our treatment of people and technology.
Gattorna’s writing gives center stage to human factors. In the review literature around Living Supply Chains, he is described as taking the reader on a journey from customers through business processes into company culture and then up to leadership, with an emphasis on how subcultures and leadership styles influence performance. His consulting practice explicitly focuses on mapping the prevailing cultures in different parts of the enterprise and matching them to the requirements of different customer segments and supply chain types. The tone is that of organizational design and change management: workshops, diagnostics, leadership coaching.
I certainly do not deny the importance of culture and leadership. If your warehouse floor is demoralized, your data is unreliable, or your procurement team is paid to chase rebates regardless of downstream consequences, no amount of clever software will save you. What I question is whether culture should be treated as the main instrument of control.
My experience has been that any process which depends on individual heroics and continuous human vigilance is fragile by design. People get tired, distracted, promoted, poached. Tacit knowledge walks out of the door every evening. If the crucial logic of your supply chain lives in a few planners’ heads or in a tangle of spreadsheets on their laptops, you might be “aligned” in the short term, but you are not robust.
That is why I put such emphasis on decision systems: software components that, given the same inputs, always produce the same recommendations, and that can be inspected, tested, and improved without persuasion campaigns. On the Lokad side, we distinguish between systems of record (the ledgers), systems of reports (the dashboards), and systems of intelligence, where the actual decisions are computed. The last category is where I believe modern supply chains are most underdeveloped. I argue more broadly for putting such decision engines at the center of the technology stack, with architectures built in service of them, in Supply Chain as Economic Bets in a Market-Driven World.
In my view, the highest‑value roles for people in supply chain are not to manually arbitrate every replenishment order, but to decide which questions the decision engines should answer, to encode the economics correctly, to design and interpret experiments, and to negotiate new options in the physical world—new suppliers, new contracts, new network configurations. Culture still matters, but primarily in whether the organization is willing to trust and continuously improve its own algorithms, or whether it prefers to hide behind dashboards and committees.
Gattorna might reply that without cultural and leadership alignment, those algorithms will never be adopted or used properly. I would agree. But I would add that if the culture work does not ultimately express itself in the code that commits capital and capacity day after day, then the transformation is incomplete. The living system must eventually leave fingerprints in the software.
Metrics, KPIs, and the problem of measurement
Gattorna’s frameworks do not ignore metrics. On the contrary, he treats performance measures as one of the levers that must be tailored to each supply chain type and behavioral segment. A lean chain will naturally emphasize cost‑to‑serve and utilization; an agile chain will accept some sacrifice on those metrics in exchange for responsiveness or innovation. The key is coherence: do the metrics encourage the behaviors and service patterns that the chosen value proposition requires?
I share the concern about coherence, but I am more skeptical of KPIs as such. Traditional supply chain control towers are filled with average service levels, forecast accuracy scores, utilization rates, on‑time‑in‑full percentages. Individually they sound reasonable; collectively they encourage gaming and local optimization. A warehouse might hit its utilization targets by resisting network changes that would increase company‑wide profit. A commercial team might push volume at all costs because their bonus depends on revenue, not margin or variability.
In my own work, I have increasingly gravitated toward measurement in terms of incremental cash uplift: if we change this policy, or deploy this new forecasting method, how much more profit did we make over a year, after risk and capital charges? This is not an easy question to answer, and you rarely get a neat controlled experiment. But making the attempt forces you to confront tradeoffs that composite KPIs tend to obscure.
Here again, the difference is not that one of us cares about performance and the other does not. It is a question of where you anchor the measurement. Gattorna anchors it in alignment with customer expectations and segment strategies; I anchor it in explicit economic impact, even if that requires more modeling effort and more humility about what we can know.
Points of convergence
It would be a mistake to exaggerate the distance between these views. On several important points, we are allies.
We both reject the idea of a single, universal supply chain configuration that can serve all customers well. Gattorna formalizes this as a portfolio of chain types matched to behavioral segments; I encounter it as a practical necessity when modeling very different products, channels, and service promises within the same company.
We both view volatility as structural, not a temporary annoyance to be planned away. He expresses this through the language of living systems, dynamic alignment, and constant reconfiguration. I express it through explicit probabilistic models and the insistence that algorithms should carry uncertainty with them rather than hiding it behind single‑number forecasts.
We both criticize the complacency of the traditional playbook: static plans, rigid S&OP cycles, mechanical KPIs, and a faith that better spreadsheets will somehow solve structural misalignments or misallocations of capital.
On these counts, I see Gattorna as an important voice pushing the field away from simple cost‑cutting and towards a more nuanced understanding of how markets and organizations actually behave. If we disagree, it is mostly about where to push next.
Where I part ways
The main place where I diverge from Gattorna is in my level of impatience with frameworks that stop at segmentation and culture.
Behavioral segmentation is a powerful lens, but it does not tell you how much inventory to place, where, when, and at what economic cost. Culture diagnostics can reveal misalignments between, say, a sales‑driven subculture and an efficiency‑driven operations team, but they do not decide which tradeoffs to make explicit in your contracts or how to price risk in your replenishment rules.
In many transformation projects I have observed from the outside, companies have embraced the language of tailored supply chains and alignment without building the mathematical and computational backbone required to act on those ideas at scale. They run workshops, print posters, rename teams. Meanwhile, the actual replenishment logic remains a tangle of safety‑stock formulas, vendor minimums, and manual overrides embedded in legacy systems and spreadsheets.
From my perspective, that is where transformation must eventually bite: in the decision logic that commits trucks, containers, warehouse space, and working capital every day. If a new strategy does not change those algorithms, you are mostly rehearsing better stories about the same behavior.
This is why I am so insistent on probabilistic models and automated decision engines. It is not because I believe algorithms are magically objective, but because they force you to write down what you believe and face the consequences. If you think a certain customer segment deserves priority in a stock‑out, encode it. If you think a certain lead‑time risk is tolerable, put that into the model and see how much it costs in expected lost sales versus capital saved. Code does not remove politics, but it crystallizes it.
In that sense, my disagreement with Gattorna is less about his high‑level diagnosis—which I broadly share—than about how far down into the machinery of the firm a supply chain theory should descend. I do not think we can stop at alignment; we have to go all the way into the algorithms.
A possible synthesis
If we are willing to stack these views rather than pick one, a kind of synthesis suggests itself.
At the top, you can use Gattorna’s dynamic alignment ideas to structure your understanding of the market. Start by observing how different customers actually buy, what they value, how they respond to variability and crisis. Group them into behaviorally coherent segments. Decide, in clear language, which segments you want to serve with continuous, collaborative arrangements, which with lean, low‑touch efficiency, which with rapid response, and which with deep, solution‑oriented flexibility. Ensure that your leadership and culture are broadly aligned with those choices.
Below that, you can apply a more decision‑centric, economics‑driven approach to the day‑to‑day operations of each chain. Within a “lean” segment, for example, you still face uncertainty about demand and lead times; you still have to decide, for every SKU in every location, how much stock to hold, when to reorder, what service‑level promises to make, how to route freight. In an “agile” or “fully flexible” segment, those uncertainties are greater and the tradeoffs sharper. The fact that the segment is labeled agile does not remove the need for quantitative discipline; if anything, it increases it.
Here, probabilistic forecasting, explicit pricing of risk, and automated decision engines can provide the necessary precision and speed. They give you a way to express, in code, the commitments implied by your alignment choices and to update them as reality pushes back.
In this layered view, Gattorna’s contribution is to help you decide which games you are playing in the marketplace and how to mobilize the enterprise around those games. My contribution is to help you place and settle the bets inside each game with greater clarity and less waste.
Closing thoughts
Supply chain is young as a formal discipline. It is unsurprising that we have multiple, partially overlapping theories of what it is and how it should work. I do not see that pluralism as a problem, provided we are clear about where each theory starts, what questions it answers well, and where its blind spots lie.
Gattorna reminds us that supply chains are not just factories and trucks and matrices in PowerPoint. They are social systems, animated by people whose behaviors and expectations matter. He is right that one‑size‑fits‑all supply chains are a costly illusion, and that culture and leadership can reinforce or sabotage whatever strategy you design.
My own work pushes in a complementary direction. I want us to be more honest and more rigorous about the economics of our choices, more explicit about uncertainty, and more ambitious in using software not just to display the state of the world but to decide how we respond to it. I have seen too many organizations stop at good intentions and new vocabularies, while leaving the engines of their supply chain—the algorithms that move goods and money—largely untouched.
If this essay has a single message, it is this: leadership, alignment, and culture are necessary, but not sufficient. So are forecasts, dashboards, and KPIs. To build supply chains that deserve the name in the twenty‑first century, we must connect the outside view of customers and markets with the inside machinery of probabilistic models and automated decisions. We must be willing to rewrite not only our slide decks, but our code.
Only then do our theories stop being metaphors and become part of the infrastructure that quietly does the work.
