Supply Chain doesn't start with Forecasts and Plans
Supply chain education usually opens with a clean map of the firm. One is shown suppliers, factories, warehouses, carriers, stores, and customers, then guided through forecasting, purchasing, production, inventory, transport, service levels, and planning cycles. The large professional bodies and the university syllabi speak a similar language. ASCM’s introductory material is organized around planning, inventory, manufacturing, distribution and logistics, processes, and procurement. CSCMP defines the field through the management of activities from product development and sourcing to production, logistics, and the information flows that coordinate them. MIT’s foundational course concentrates on forecasting, inventory, and transportation. SCOR, meanwhile, expresses the subject through standard process families and gives planning an explicit central role. Standard textbooks follow the same road.
In Introduction to Supply Chain, I chose another opening. I begin with decisions, scarce resources, and uncertainty, then move through economics, information, intelligence, engineering, deployment, and the reasons the field remains intellectually stuck. That sequence was deliberate. Before a supply chain can be planned, it has to be understood. Before it can be understood, one has to ask what sort of object it is. My answer is that it is not, in the first instance, a chart of functions. It is a domain of economic choice governing the flow of physical goods.
The mainstream opening
The conventional syllabus has an obvious strength: it mirrors the visible organization of firms. Departments exist. Warehouses exist. Procurement teams exist. Transport budgets, master schedules, demand plans, and inventory targets all exist. A course that proceeds through those objects feels concrete. It gives students vocabulary quickly, and it fits certification formats very well. That is one reason ASCM can teach the subject as a progression through modules such as planning, inventory management, distribution and logistics, manufacturing, and purchasing, and why CSCMP can package foundational training into tracks like demand planning, inventory management, procurement, warehousing, transportation, and manufacturing operations.
This structure also elevates planning very early. ASCM’s planning certificate centers on synchronized planning, S&OP, master scheduling, MRP, purchasing and production control, and distribution planning. SCOR describes planning as the work of determining requirements, balancing them against resources, and deciding how to close the gaps. The textbook tradition does much the same by moving from strategic fit and supply chain drivers into demand forecasting, aggregate planning, S&OP, inventory models, transport, sourcing, and pricing. The student is invited to see the supply chain as a large coordinated apparatus whose success depends on better forecasts, better plans, and better compliance with those plans.
I do not dispute the importance of those topics. I dispute their place in the order of learning. When forecasting and planning come first, they inherit an authority they have not earned. The student learns to treat the plan as the natural center of the field, while service levels, forecast accuracy, and similar measures appear as objective guides. Yet these objects are secondary. They matter only insofar as they help make better commitments under uncertainty. They do not define the discipline. They arise from a deeper problem, and that deeper problem is economic before it is procedural.
The omitted first principle
I start with the commitment itself. A purchase order ties up cash in one item rather than another. An allocation grants service to one channel and withholds it from another. A truck dispatched here is no longer available there. A price cut changes the expected path of inventory just as surely as a stock transfer does. This is why I place pricing inside the remit of supply chain rather than at its edge. Every deliberate action that changes what moves, when it moves, where it moves, or how much moves, is a supply chain decision. Once seen from that angle, the field sits inside economics before it sits inside process management.
That shift in viewpoint changes the status of familiar metrics. Service level, forecast accuracy, utilization, and adherence to plan can be acceptable instruments inside a calculation, much as laboratory glassware is acceptable inside chemistry. Trouble begins when the instrument begins to rule the organization. A service target has no economic virtue on its own. A forecast can grow more accurate in the narrow statistical sense while the firm ties up more capital, accumulates more markdown exposure, and becomes slower to adapt. A good decision is not one that flatters a dashboard. It is one that makes better use of scarce resources than the alternatives that were available at the time.
Uncertainty also takes a different place. Mainstream academia does not ignore it. MIT, for example, teaches analytic tools for forecasting, inventory, and transport from the start. My disagreement concerns the role uncertainty is allowed to play. In the conventional order, uncertainty appears as a complication added to planning. I place it nearer the entrance because it gives the whole subject its character. Supply chain data are often sparse, lumpy, batched, and exposed to abrupt shifts in demand, lead times, returns, substitutions, and prices. A single-number forecast describes such situations badly. The meaningful object is a range of plausible futures and the economic consequence of acting under each one.
This is why the familiar safety-stock recipe makes a poor foundation. It treats items as if they did not compete for the same pool of working capital. It asks managers to pick service targets whose link to profit is assumed rather than demonstrated. It relies, in its standard form, on statistical assumptions that are convenient on paper and frail in the records of actual firms. One can keep repairing the formula, but the more serious the repair becomes, the less remains of the elegant classroom object that made the method attractive in the first place.
The profession that follows
Once supply chain is introduced as economic choice under uncertainty, the image of the practitioner changes as well. The mainstream curriculum often points toward a planner or analyst who coordinates functions, tends the forecasts, and improves performance metrics with the help of software. I have in mind a practitioner who designs the rules by which the firm turns records into commitments. That work requires a grasp of the meaning of the firm’s data, the shape of its uncertainties, the trade-offs hidden in its operations, and the money attached to each of them. It is judged by the quality of the decisions that reach the ledgers: purchase orders issued, transfers booked, prices adjusted, stock positioned, capacity committed.
From there, software moves to the center of the subject. A large supply chain requires too many daily decisions for clerical supervision to remain a respectable operating model. Firms need software that consumes authoritative records, carries the trade-offs in money terms, and writes back auditable decisions. That is why I devote so much space to information, intelligence, engineering, and deployment. Today, the difficult work is to encode sound judgment over the flow of goods.
This, in turn, explains the divergence from much of today’s courseware. Many published programs still revolve around point forecasting, synchronized planning, planning modules, inventory targets, service levels, and the routine administration of packaged systems. Even where mathematics is present, it is often taught outside an explicit economic logic. Programming, data engineering, and the semantic fidelity of business records receive far less attention than they now deserve. The graduate is prepared to supervise the machinery of planning. Less often, that graduate is prepared to author the machinery of decision.
I understand why the conventional introduction persists. It is easy to standardize, easy to certify, and easy to align with existing departments and software categories. It provides a common language across functions. It also allows the hardest questions to wait outside the classroom a little longer: which commitments deserve scarce capital, which forms of flexibility are worth paying for, which uncertainties matter enough to price explicitly, and which decisions should be automated rather than merely reported. Those questions are less convenient than a process map. They are also closer to the daily truth of the work.
Where a discipline begins matters. Start with processes, plans, and local metrics, and one tends to produce practitioners who manage those objects more carefully. Start with scarce resources, rival commitments, and uncertainty, and one gives practitioners a better chance of making sound economic choices and of building the software that can execute them at scale. Once the starting point shifts, the rest of the subject settles into a different order.
