Framing the engagement and decision rights

When a sponsor says, “Our supply chain is slow,” the statement has not yet defined a testable problem. The sponsor may mean elapsed time from accepted order to delivery, delayed material purchases, long production time, or poor updates about an order that is otherwise on schedule. The consultant first specifies the unit of analysis: which orders, over what period, and between which start and end events. Pooling domestic standard web orders with customized export orders can make an average improve while service to one segment deteriorates. “Reduce delivery time” becomes a useful outcome only after the clock, cohort and team's responsibility are fixed.

ASCM's SCOR Digital Standard helps name the broad processes: Plan, Order, Source, Transform, Fulfill and Return, along with supporting orchestration. It is a map for questions, not proof of a cause. If delay occurs between order confirmation and inventory allocation, begin with the promise and allocation decision, then trace its relation to planning and sourcing. If the shipment leaves on time but arrives late, investigate transport and downstream event data. A process label does not justify buying a system or replacing a supplier. The reference describes process categories; the organization's records must establish where time was actually lost.

A practical evidence register has four columns: claim, observation, assumption and test needed. “One third of orders are late” is a claim. A system report showing 120 of 360 orders beyond a promise date is initial evidence. Whether timestamps are valid and the order population complete remains an assumption until checked. A warehouse manager may say the supplier caused the delays; that is an explanation, not a fact merely because the manager is experienced. Test it by comparing material receipt to pick release dates, including both late and on-time orders. Record each table's source, extraction date and definition owner because a corrected definition may change the apparent rate.

A good engagement question connects the problem to a decision that can be made: “Over four weeks, determine why standard direct-channel orders miss their promise date, compare changing the promise rule with changing allocation, and estimate effects on service, cost and working capital.” That is narrower than “fix the supply chain” and broader than a preselected cause. State exclusions as well: long-term supplier contracts, export orders and technology changes that need information-security approval. Evidence may later show that an exclusion blocks a useful recommendation. In that case, request an explicit scope change with reasons rather than expanding the work silently.

Decision rights affect the usefulness of the analysis. An analyst can calculate delays and inspect samples; a customer-service owner can propose another promise; sales may challenge its effect on conversion; and the authorized sponsor chooses the acceptable trade-off. Distinguish who provides data, who validates it, who bears the decision's consequences and who approves. Seniority does not establish the scientific cause. Equally, a correct formula does not make a proposed action executable if no one in scope can authorize it. The opening memo should name one decision, a success measure, an approver and a condition that would invalidate the conclusion.

When stakeholders disagree, do not reconcile them with a vague promise. Sales may prefer a shorter quoted lead time to improve conversion, operations a promise it can meet, and finance lower inventory. Ask the sponsor to rank decision criteria before seeing results: promise adherence, margin after expediting, and working capital, for example. Name who may change the promise rule and who may fund extra stock; they may differ. If the commercially attractive option breaches a cash constraint, show that trade-off rather than hide it in one service average. Define a scope-reopening trigger too. If most late cases prove to be excluded export orders, do not generalize a domestic sample to exports.