Norwich Cold Chain: Depot, 3PL or Absorb the Peak?
Norwich Cold Chain is a regional refrigerated distributor with GBP 42m revenue and a single depot that reaches 88% utilisation in peak season. The board is choosing a three-year capacity strategy from three options: (A) build a second depot - GBP 4.2m capex, 15-month build, GBP 0.6m annual fixed cost; (B) contract a national 3PL - variable cost roughly 11% higher per pallet, no capex, 60-day mobilisation; (C) absorb peak with overtime and rented trailers. One grocery customer supplies 38% of volume; its contract expires in 18 months and its procurement team has signalled it will tender. NCC has already spent GBP 250k on depot design and planning permission, which the operations director cites as a reason to proceed. The finance director notes the balance sheet can support option A but covenant headroom falls to 1.8x. The sales director argues that 'we have never lost a customer and never will'. No demand forecast has been formally documented.
1. Identify the decision-relevant costs and the reasoning errors in the positions of the operations, finance and sales directors. State explicitly what should be excluded from the analysis.
Read the worked answer
Exclude: the GBP 250k already spent on design and permitting is sunk. It cannot be a reason to proceed. What is not sunk is the option value of an already-permitted site - it may shorten option A's timeline by several months. That is a timing benefit to be valued, not a justification.
Operations director - sunk-cost fallacy and confirmation bias. He is defending a past decision rather than comparing future cash flows.
Sales director - availability bias and an untested claim. 'We have never lost a customer' is a statement about the past, drawn from a small sample, in a market where the anchor's procurement team has already announced a tender. It also ignores that a 38% concentration means the contract, not the depot, is the dominant variable.
Finance director - a constraint presented as a conclusion. Covenant headroom of 1.8x is real, but without a probability of breach under plausible demand scenarios it cannot rank the options.
Also missing: option C is not 'do nothing'. It has overtime, trailer rental, service-failure and penalty costs, plus a revenue-at-risk estimate if service levels slip in peak.
What belongs in the analysis: incremental three-year cash flows per option, including exit costs and contract break clauses; cost per pallet at several volume levels; the value of reversibility (3PL can be exited, a depot cannot); revenue at risk if the anchor leaves; and covenant headroom under a stress case. Every option should be valued conditional on the anchor renewing and on it departing.
2. Recommend a preferred option and the conditions under which you would switch to another. Express these as explicit triggers.
Read the worked answer
Recommendation: option B (3PL) as the base case, with the permitted site option kept alive.
Reasoning: the dominant uncertainty is the 38% anchor contract, which resolves in 18 months. Option A commits GBP 4.2m of irreversible capital and GBP 0.6m of annual fixed cost 15 months before that uncertainty resolves, and it does so on an undocumented demand forecast. Option B costs 11% more per pallet but preserves the ability to exit, which is precisely what an unresolved anchor contract demands. It also keeps covenant headroom intact for whatever the tender outcome requires.
Switch triggers to option A (any two together justify re-opening the decision):
- The anchor renews on a term of three years or more with a volume commitment.
- A second contracted customer covers at least 15% of forecast volume.
- Peak utilisation exceeds 92% for two consecutive seasons under option B.
- 3PL variable cost rises above roughly 18% premium, eroding the flexibility premium.
- The permitted site's consent is due to lapse, converting a real option into a deadline.
Switch triggers to option C: only if the anchor contract is lost or renegotiated down by more than 20% of volume, in which case capacity shrinks and capital deployment should be paused entirely.
Key discipline: option B must be agreed with a documented exit right and a defined mobilisation service level, otherwise the 'reversible' option is reversible in theory only.
3. The CEO wants to announce the decision at the next board meeting in three weeks. What should happen before the decision, and how should the anchor-contract uncertainty be handled in the board paper?
Read the worked answer
Before the meeting:
- Produce a documented demand forecast with three scenarios (anchor renews, anchor re-tenders and stays at reduced volume, anchor lost) and attach probabilities with their basis. No forecast should be accepted from a single function.
- Build three-year NPVs for A, B and C under each scenario, plus a probability-weighted expected value. Include exit costs, mobilisation costs and revenue at risk - not just operating cost per pallet.
- Stress-test covenant headroom under option A in the anchor-lost case. If headroom breaches, that is close to a decisive finding.
- Obtain legal advice on the 3PL break clauses and on the expiry terms of the planning permission, so the option value of the site is quantified rather than asserted.
- Have the board minute the treatment of the GBP 250k as sunk.
Handling the uncertainty in the paper: present the decision as a choice under uncertainty, not a single-point forecast. Show the regret table (which option hurts most if the anchor leaves), state the switch triggers explicitly, and ask the board to approve option B plus a standing authorisation to re-open option A when the triggers are met.
On timing: three weeks is a board-calendar deadline, not a decision deadline. The genuine timing constraint is 3PL mobilisation at 60 days and the anchor tender timetable. If the analysis cannot be completed properly, the board should approve a conditional direction and a date for the final decision - announcing a number to fill a slot is how bad capital decisions get made.