Two Component Suppliers: Effective Cost, Volume Commitment and a Sensitivity Check
A manufacturer needs 48,000 usable units of a machined component next year. Supplier A quotes $2.40 per unit with no commitment and a 1.8% defect rate; defective units are scrapped and cannot be reworked. Freight from A is $0.05 per unit shipped. Supplier B quotes $2.15 per unit against an annual committed volume of 42,000 units, with any units beyond the commitment billed at $2.35; B's defect rate is 3.4%, but B's defective units can be reworked at $0.60 each and returned to usable stock. Freight from B is $0.09 per unit. B ships in larger lots, adding $1,800 per year in holding costs. Both suppliers have equal lead times and equivalent quality certification, and the component is single-sourced either way. The purchasing manager says B is obviously cheaper because $2.15 is 25 cents below A's price, and recommends signing the commitment today.
1. Compute the annual cost of each option on a per-usable-unit basis and make a recommendation. Does the manager's reasoning hold?
Read the worked answer
Supplier B is cheaper by about $8,260 a year - but not for the reason the manager gives.
Supplier A - buy enough to cover scrap:
- Units to order = 48,000 ÷ (1 − 0.018) = 48,880
- Landed cost = 48,880 × ($2.40 + $0.05) = $119,756
- Effective cost per usable unit = $2.495
Supplier B - order 48,000, rework the defects:
- Purchase = 42,000 × $2.15 + 6,000 × $2.35 = $90,300 + $14,100 = $104,400
- Freight = 48,000 × $0.09 = $4,320
- Rework = 48,000 × 0.034 × $0.60 = $979.20
- Holding = $1,800
- Total = $111,499
- Effective cost per usable unit = $2.323
B wins by $8,257, roughly 7%.
The manager's reasoning is wrong even though the conclusion is right. The per-unit sticker gap is not the driver: B's defect rate is nearly double A's and its freight is four cents higher. What decides the comparison is that reworking a B defect costs $0.60, while covering an A defect costs $2.45 of scrapped material plus freight.
2. How far could B's above-commitment price rise before Supplier A becomes cheaper, all else equal? What does that tell you about the robustness of the decision?
Read the worked answer
The threshold is about $3.73 per unit on the 6,000 overage units.
Set B's total equal to A's and solve for the overage price p:
90,300 + 4,320 + 1,800 + 979.20 + 6,000p = 119,756 97,399.20 + 6,000p = 119,756 6,000p = 22,356.80 p ≈ $3.726
So B remains cheaper unless its overage price exceeds roughly $3.73 - a 59% premium over the committed rate. The cost ranking is therefore not fragile to a small pricing change; the margin absorbs a large one.
The decision is fragile in a different dimension: volume. The 42,000-unit commitment is a floor. If demand falls to, say, 38,400 units, B is still paid for 42,000. At exactly 42,000, B's cost is (90,300 + 3,780 + 1,800 + 856.80) ÷ 42,000 ≈ $2.30 per usable unit - still below A - so the commitment is not expensive on its own; it becomes expensive only if the forecast is materially wrong, which is a demand-risk question rather than a unit-price question.
3. What non-cost consideration could reasonably overturn the recommendation?
Read the worked answer
Three, in descending order of force.
Downside volume risk. The commitment is a floor, not a target. A 20% demand shortfall means paying for 42,000 units while consuming 38,400 - roughly $8,400 of unusable material at B's blended rate, which wipes out essentially the entire $8,257 advantage. The recommendation rests on a forecast being right, and that risk is not priced in the comparison.
Quality escapes. The model assumes every defect is caught and fully restored at $0.60. If inspection is sampled rather than 100%, a 3.4% defect rate leaks bad units into assembly, and the cost of a line stoppage or a customer return is far above $0.60 per unit. A's higher price partly buys a lower incoming defect rate.
Supply concentration. The comparison treats both options as single-source, but A at no commitment is a cheap option to keep qualified. Losing that second source to save $8,257 a year is a trade the cost model does not evaluate and management should price explicitly.