Kestrel — stages 1 and 2

Staged, not shown

This lesson and the next are the group PSP case in rehearsal. You write each stage as an exam, commit it, and only then read the model with its rubric lines labelled — the discipline that makes the model useful rather than something to copy. The prose below says what the stage must contain for this case, and gives you the one calculation stage 2 needs; it does not give the answers.

Stage 1 for Kestrel — what R-psp-1 needs here

Role and task, from L11.1: junior analyst, finance; a memo to the CFO recommending which one of five apparel suppliers to drop, by 27 February. Context in two or three facts: three stores and online; five apparel suppliers with purchases from $150,000 to $520,000; consolidation to four is expected to save about $22,000 a year of relationship administration and earn volume discounts. Stakeholders and what each wants — this is where Kestrel is richer than Lakeshore: the CFO (the saving, and a decision he can defend); merchandising, in the person of Lena Bramwell (a workable assortment — and, as the case planted, an interest in Cedar & Co.); the store staff who handle receiving and returns (fewer defects, reliable deliveries); the customers, including those who ask for Grey Jay's Canadian-made line by name; the five suppliers, one of whom will lose the account; and — the stakeholder most stage-1 responses forget — the workers in the offshore factories that make Tundra Line's goods, whose conditions the failed labour audit is about. Constraints: the 27 February deadline and the March order cycle; the $22,000 saving per relationship as the financial objective; Northwind's three-year contract, which has a six-month exit notice, so it could not be dropped before the fall season even if the analysis pointed there; information — one year of data, one audit each. The kind of analysis and why: a weighted decision matrix of the five suppliers on criteria drawn from the exhibit and the stakeholders, with a margin-dollar calculation, because the question is a ranked choice among five on several objectives and the CFO expects numbers.

Stage 2 for Kestrel — what R-psp-2 needs here

Issues, each in one sentence with a fact, problem separated from cause and symptom. The main issue is the task: which supplier is weakest on Kestrel's objectives? Around it, the issues the exhibit raises: Tundra Line — the largest account and the highest margin — failed its labour audit, and remediation is only "pending": a reputational and ethical risk that a rate-and-volume view misses. Lena Bramwell's shortlist recommends dropping Grey Jay while her sister owns Cedar & Co.: a conflict of interest that makes the shortlist's recommendation unreliable — not a finding about Lena's honesty, a fact about the source. Sable Peak's 3.8 % defect rate comes from one shipment: a small-sample figure that should not be read as a rate without asking what the rate is with that lot excluded. Cedar & Co. is last on both quality (4.5 %) and reliability (82 %), and is a small account. Ranked by impact on the objectives: the choice of supplier first (it is the task); the audit failure second (the largest reputational exposure); the conflict third (it decides whether the shortlist can be trusted); the small sample last (it changes one score). Relevance to stakeholder goals stated for each. Assumptions written down: that the volume of a dropped supplier can be re-sourced to the survivors at similar or better margin; that the $22,000 is real and per relationship; that one year of data is representative.

The calculation stage 2 should make

The exhibit gives purchases and margin rate. The CFO thinks in margin dollars — what each supplier's goods earn Kestrel in a year, and therefore how much is at stake if a supplier is dropped and its volume is not fully re-sourced. Sales on a supplier's goods are purchases ÷ (1 − margin), and margin dollars are sales − purchases. The callout gives the five figures. Notice that the two smallest are Cedar & Co. and Grey Jay, at about $104,000 and $110,000 — close — and that only one of the two has a quality and a reliability problem. That is the quantitative fact Lena's note does not mention.

Now write both stages

Stage 1 first, then stage 2, each against its rubric. Commit each before reading its model.