The Bracket Supply Base After a Nine-Day Line Stop

Step 1 - Annual spend by supplier

Supplier	Units	Price	Spend	PPM
A	148,800	$14.20	$2,112,960	1,200
B	48,000	$15.05	$722,400	300
C	28,800	$13.80	$397,440	2,900
D	14,400	$16.40	$236,160	450
Total	240,000		$3,468,960

Weighted average price = $3,468,960 ÷ 240,000 = $14.45.

Step 2 - What price-only logic picks

The two lowest unit prices are C ($13.80) and A ($14.20). Consolidating to A+C would move C from 12% to roughly 30% of volume - a 2.5× increase in the volume flowing through the worst quality performer.

Step 3 - Add the containment cost that is already measurable

Containment cost per part = (PPM ÷ 1,000,000) × $45.

• A: 0.0012 × $45 = $0.054 → $14.254 quality-adjusted
• B: 0.0003 × $45 = $0.014 → $15.064
• C: 0.0029 × $45 = $0.131 → $13.931
• D: 0.00045 × $45 = $0.020 → $16.420

C is still nominally cheapest - which is exactly the trap. The $45 figure captures only containment labor; it prices a rejected lot, not a rejected lot that reaches a customer or stops an assembly line.

Step 4 - What the analysis actually misses

1. Concentration risk. A is already at 62%. Adding to it, or relying on a two-supplier base where one is A, recreates the single-point failure that just cost nine days of output.
2. Capability, not just PPM. C's 2,900 PPM at low volume suggests the process is not in control; scaling volume scales the defect stream. PPM measured at 12% of volume is a small sample and a poor predictor of capability at 30%.
3. Tooling lock-in. Two variants are only sourceable from A. No consolidation is real until that is resolved.
4. The "volume discount" is unquantified. No supplier has offered a price per volume band. The manager is trading a known cost for an assumed benefit.

Verdict: the price-only answer is A + C. It is the wrong answer, and it is wrong on information already in the file.

Recommended structure: A and B as the strategic pair, with a defined exit/development decision on C and D.

Why B must grow despite the price premium

B costs $0.85/part more than A. Moving 24,000 units from A to B costs 24,000 × $0.85 = $20,400 a year in price, and returns:

• A's share drops from 62% to 52%, which materially reduces the single-point-of-failure exposure that caused the nine-day stop.
• The defect stream falls: 24,000 units at A's 1,200 PPM = ~29 defects/yr; the same units at B's 300 PPM = ~7 defects/yr. At $45 containment that is only ~$990 saved - the case rests on continuity, not containment.
• B is the only supplier whose quality performance is consistent with a safety-critical medical device.

Proposed split

• A: 50–52%. Still the volume leader because it is the lowest-cost capable supplier, but capped.
• B: 33–35%. Grown to a genuine second source with capacity committed in writing.
• C: hold at ≤12% or exit. Set a gate: 800 PPM and Cpk ≥ 1.33 on the critical bore within two quarters, or volume is removed. Do not increase C's share as a reward for price.
• D: exit or retain only as overflow for the third variant. 6% of volume is too small to manage, and the $16.40 price is not defensible unless D is the only source for a variant.

Transition sequencing

1. Resolve tooling ownership first - nothing else can move until the two A-only variants have transferable or duplicate tooling.
2. Qualify B on the two A-only variants (first-article, capability study, pilot lot).
3. Ramp B in 5-point increments over two quarters so no supplier sees a step change it cannot absorb.
4. Issue C a written improvement gate with a dated consequence. A gate without a consequence is a suggestion.
5. Put a dual-source requirement into the commodity strategy for every variant, not just the two that happen to be exposed today.

Cost of the plan: roughly $20k–$25k a year in incremental piece price plus tooling duplication, against a nine-day line stop that has already been paid for once.

The exposure

Northline's ability to move two of three variants depends entirely on A's cooperation. There is no documented ownership, no obligation to return or transfer, no access to drawings, and no provision for A's insolvency or a dispute. In practical terms A holds a hostage, whether or not it intends to.

Contractual actions

1. Establish ownership in writing. The tooling was presumably paid for by Northline - confirm with the original PO or capital authorization. If Northline paid, ownership is Northline's and the agreement should say so explicitly, with a marking requirement (nameplate/asset tag on the tool).
2. Right to possession and transfer. Add an obligation to return tooling on demand within a stated period (e.g., 30 days), at Northline's cost, in serviceable condition, plus the right to enter and remove it.
3. Drawings and documentation. Require delivery of native CAD, tool drawings, maintenance records, and a documented tool condition report at transfer. This is cheap now and expensive later.
4. Insurance and care. Require A to insure the tool at replacement value with Northline named, and to maintain it to the OEM's schedule.
5. Continuity clause. Require a transition-assistance period (e.g., 90 days of support to a new source) triggered by termination or by failure to meet quality gates.

Operational actions

• Pull the drawings off A's server now and store them in Northline's PLM; do not wait for the contract amendment.
• Commission a duplicate tool set for the two exposed variants. Treat the cost as insurance and amortize it against the volume; the break-even is one avoided line stop.
• Do not consolidate volume onto A until both actions are complete. Any consolidation done before tooling is secured increases dependence on the one supplier with the worst continuity record.