One supplier or two? Single vs dual sourcing when the part is critical

1. Before you start

Two words decide how a factory leans on its suppliers, and they pull in opposite directions. Single sourcing means buying a part from one supplier on purpose — to concentrate your volume and win a lower price, a tighter quality bar, and a deeper relationship. Dual sourcing (or multi-sourcing) means qualifying a second supplier for the same part, so that if one stops, the line keeps moving. A tiny example. Buy 100,000 identical connectors a year: give all of it to one supplier and they cut the price 5% for the volume; split it across two and you lose most of that discount and pay to qualify the second plant — but a fire at one supplier no longer stops your whole line. The first buys price; the second buys insurance. The skill is knowing which one the part in front of you actually needs.

Three honest statements before you start:

This is a Decide course. You read a situation, learn the ideas, and make a call. You do not write or run any code, and there is no calculator here — the whole task is judgment.
The company in the case, Meridian Devices, and the firm in your final call, Kestrel Robotics, are composite — invented from ordinary, realistic dynamics so the reasoning is clean. No number here is a claim about any real company.
This is not a certification. It proves, to you, that you can weigh a supplier’s price against the cost of the day it fails to deliver — and defend the call.

The difficulty is the trade-off: the cheaper answer and the safer answer are usually different suppliers, and the report in front of you flatters the cheaper one.

2. The Situation

Meridian Devices, a composite maker of connected home-energy monitors, buys one part it cannot build around: a specialized power-management chip at the heart of every unit, from a single supplier in one industrial park. Procurement can consolidate all of next year’s volume with that supplier for a better price — or spend real money qualifying a second supplier in another region as a backup. The head of operations must choose, and the choice is not really about price: it is about what a six-week outage would cost if the one supplier ever went dark.

The trap is that the single supplier looks best on every line of the cost sheet, because the cost of the outage that has not happened yet appears on no line at all.

3. What you’ll be able to do

After this course you will be able to:

State the core trade-off — the price and quality a single source buys versus the disruption risk it concentrates — and say why the cost sheet hides half of it.
Reason about the expected cost of a disruption qualitatively, as probability times impact, and see why a rare-but-severe event can outweigh a certain, visible discount.
Name the hidden costs of dual sourcing (lost volume price, qualification, complexity and quality variance) so you never treat a second supplier as free insurance.
Use the Kraljic lens to match sourcing to the part — commodity versus strategic, low versus high exposure — and defend a single- or dual-source call, naming the one fact that would flip it.

4. Prerequisites & time box

Prerequisites: none beyond comfort reading a small table of numbers and the idea that a supplier is a point of failure as well as a price. No spreadsheet or setup — the Decide hall is read-and-decide in the browser; see the Decide hall’s how-to-read page if this is your first concept course. No prior Decide course is assumed.

Time box: about 24 minutes of reading (measured), plus real thinking time on the call in section 7. That is under the 25-minute cap for a concept course.

Difficulty: 6 / 8 — a manager-level decision: several factors move at once and the naive reading points the wrong way, so you have to reason past it.

Free-tier honesty: no signups, no paid tools, requires_gpu: false. Nothing here costs money to learn.

5. The case & where the numbers come from

Meridian Devices is a composite electronics company: its volumes, prices, and disruption estimates are in-course illustrative assumptions, chosen for clean reasoning and clearly labelled as such — not drawn from or claimed about any real firm. The frameworks — the Kraljic purchasing portfolio, second-sourcing in electronics, supply-chain risk as probability times impact, total cost of ownership — are standard and cited in section 11. Every comparison below is done inside the course from these figures, so you can reproduce each by hand.

The part in question, as the cost sheet reports it (all figures illustrative):

Item	Figure (illustrative)
Annual volume of the power-management chip	500,000 units
Single-source price if all volume is consolidated	$3.80 / unit
Contribution lost per finished monitor if a unit can’t ship	$40
Weeks of production a serious supplier outage would cost	~6 weeks

Hold those four numbers. The cost sheet shows only the first two. Section 6 is about the two the cost sheet leaves off — what an outage would cost, and how likely it is — because that is where the sourcing decision actually lives.

6. The Concepts

The core trade-off

Every sourcing decision balances two goods in tension. Concentrating volume with one supplier buys real things: a lower unit price (you are its biggest customer), more consistent quality (one process, one qualification, one set of tooling), and a deeper relationship (priority when parts are scarce). Meridian’s $3.80 price exists because it promised all 500,000 units to one plant. Spreading volume across two or more suppliers buys a different thing: resilience. If one plant floods, loses power, gets allocated out, or simply fails a lot, the second keeps your line running.

The two goods trade against each other. Split the volume and you are no longer anyone’s biggest customer, so the price rises and the discounts thin. Keep it whole and you are efficient right up until the day your one supplier cannot ship — and then you are not efficient, you are stopped. A single supplier is a single point of failure: one event, and the whole flow of that part halts. Price and consistency sit on one side, exposure to a line-down event on the other, and the reason it is hard is that the first side is on the cost sheet and the second is not.

Expected cost of disruption

To compare a certain discount against an uncertain disaster, you have to put them in the same units. The tool is the oldest one in risk: expected cost = probability × impact. You do not need a precise number — you need an honest order of magnitude, so a rare-but-severe risk stops hiding behind the word “unlikely.”

Work Meridian’s exposure, all figures illustrative. If the one supplier went dark for 6 weeks with no backup, Meridian could not build for six weeks — roughly 6/52 of the year, so about 58,000 units of lost production; at $40 of contribution per monitor, that is roughly $2.3 million of margin gone, before counting empty retail shelves and lost trust. Now the probability: a plant in one industrial park faces fire, flood, quake, power loss, or being allocated out in a shortage. Put the chance of a serious multi-week outage in a given year at, illustratively, about 15% — once in six or seven years. Then the expected cost of staying single-source is about 0.15 × $2.3M ≈ $345,000 a year.

That number is the point: not on any cost sheet, not certain, and not small. A 15% chance of a $2.3M event is not “unlikely, so ignore it” — carried year after year it is a six-figure cost you pay in risk whether or not the bad year has arrived. The discipline is to size it, not feel it: probability times impact turns “it probably won’t happen” into a number you can set beside the discount.

The limits, so you stay honest. The 15% and the $2.3M are estimates, and reasonable people will argue them — that argument is the work. A second source does not drive the expected cost to zero either: if both suppliers sit in the same region, a regional event takes them both (their risks are correlated), and the backup you paid for fails when you need it. Expected cost is a lens for comparing options honestly, not a promise about any single year.

The hidden costs of dual sourcing

If a second supplier buys down a $345k risk, why not add one to everything? Because dual sourcing is never free, and its costs hide as easily as the disruption cost does — just on the other side of the ledger.

You lose the volume price. Split 500,000 units 70/30 and neither supplier gets the whole prize, so the consolidation discount thins. If the blended price drifts from $3.80 to about $4.05, that is $0.25 × 500,000 ≈ $125,000 a year, every year, certain.
Qualification is real money and time. A second supplier for a critical chip must be audited, sampled, and validated before it ships a single production unit — tooling, test runs, engineering hours. Illustratively $250,000 one-time; over three years, roughly $85,000 a year.
Complexity compounds. Two suppliers means two relationships, two quality systems, two incoming inspections, and a new failure mode — parts meant to be identical that drift apart, so a defect traces back to “which source?” Call the ongoing overhead a further ~$40,000 a year.

Add them and dual sourcing costs Meridian on the order of $250,000 a year — certain, visible, and paid whether or not a disruption comes. That is the honest price of the insurance. It sits below the ~$345k expected disruption cost, which makes dual sourcing defensible for this part — but the two numbers are close enough that the call is a judgment, and a different probability or a smaller impact would flip it. Treat a second source as free safety and you will over-insure cheap parts and quietly destroy margin.

The Kraljic lens

You cannot run the expected-cost argument on all 4,000 parts in a product. You need a way to sort them, and the standard one is the Kraljic matrix — a purchasing portfolio that places each part on two axes: profit impact (how much this part’s cost and availability matter to the business) and supply risk (how hard it is to get, how few suppliers exist, how exposed they are). The two axes make four quadrants, and each wants a different sourcing posture:

Strategic (high impact, high supply risk) — Meridian’s chip. Manage the relationship and the risk, not just the price: dual or regional sourcing, deep partnership, buffer stock. This is where a second source usually earns its cost.
Leverage (high impact, low supply risk) — matters to cost but has many capable suppliers. Use competition: single-source to the best price, re-bid often, switch freely, because switching is easy and the risk is low.
Bottleneck (low impact, high supply risk) — cheap but scarce or single-origin. Small money, but it can stop the line; assure supply with stock or a backup even though it barely shows in spend.
Non-critical (low impact, low supply risk) — commodity fasteners and resistors. Minimize effort; single-source or catalogue-buy and move on.

The lens keeps you from the same mistake in both directions: over-managing a commodity, and under-managing the one strategic part that can halt everything. A part’s quadrant, not a reflex, sets the strategy — and the power-management chip sits squarely in strategic, which is why Meridian is even considering a second source.

When each is right

Put the three ideas together into a working rule.

Single-source when the price prize is real and the disruption exposure is low — a leverage or non-critical part with many suppliers, low switching cost, and no catastrophic line-down if one slips. Concentrate the volume, take the price, and re-bid to keep the supplier honest.
Dual- or multi-source when the impact of losing the part is high and the exposure is real — a strategic or bottleneck part, few suppliers, a long qualification lead time, or a supply base concentrated in one region or plant. Here the expected disruption cost dominates, and paying the split-volume premium is buying insurance that, over time, pays.
Whichever you choose, make the hidden cost explicit. Single-sourcing has an unpriced risk cost; dual-sourcing has an unpriced complexity and price cost. The bad decision is not “single” or “dual” — it is choosing either while pretending its hidden cost is zero.

For Meridian’s chip, the reasoning points to a second source: the part is strategic, an outage costs about $2.3M, the exposure is a single plant in a single park, and the ~$250k annual cost of a backup sits below the ~$345k expected cost of carrying the risk alone. But the honest version names what would flip it — a much lower outage probability, a cheaper way to cover the impact (a bonded buffer stock), or a second source so close to the first that their risks are correlated and the insurance is hollow.

7. Your Call

You have seen how the core trade-off, the expected cost of disruption, the hidden costs of a second source, and the Kraljic lens decide Meridian’s chip. Now a different decision lands on your desk.

Kestrel Robotics is a composite maker of warehouse robots. Its robots depend on one single-sourced part: a specialized vision-sensor module, bought entirely from one supplier whose only plant sits in a single region. Two things just changed at once. A new import tariff has been announced on components from that region, due next year — a policy shock, not a plant failure. And that same supplier, sensing Kestrel might look elsewhere, has offered an exclusive deeper discount: sign for all of next year’s volume and stay single-source, and the price drops further. Kestrel’s finance lead wants to take the discount and lock it in. Operations is uneasy: the discount is certain, but so is the fact that one tariff or one regional event now sits between Kestrel and every robot it ships.

How this differs from the taught case (the transfer): this is a different company and sector (Kestrel Robotics in warehouse robotics, not Meridian’s home-energy electronics); the figures are different, so the reasoning must be redone rather than recalled; it is a different kind of decision (respond to a supplier’s exclusive price offer under a looming policy shock, not decide from scratch whether to qualify a backup); and it adds a constraint the taught case did not have (an announced tariff and a supply base concentrated in one region, tightening the incumbent’s grip just as its risk rises). The core concept is the same: single versus dual sourcing risk — the price a single source buys against the disruption it concentrates.

8. Self-check

Before you write the memo, make sure you can say each of these in one line:

What two goods do a single source and a second source each buy, and why can’t you have both?
How do you size a disruption you have never seen — the two numbers — and why does a low probability not make a line-stopping impact safe to ignore?
What are the three hidden costs of a second supplier, and why do they mean you should not dual-source everything?
Which Kraljic quadrant is the part in, and what one fact — a probability, an impact, or a cheaper way to cover the risk — would flip your call?

If any is fuzzy, reread section 6: the core trade-off, the expected cost of disruption, the hidden costs of a second source, and the Kraljic lens are the whole course.

9. Stretch

Push the thinking further on your own:

Meridian’s second source only helps if its risk is not tied to the first’s. Sketch what “independent” must mean here — region, power grid, sub-tier suppliers — and describe two suppliers that look separate on paper but share a hidden single point of failure.
Instead of a full second source, Kestrel could hold a buffer stock of modules to ride out a disruption. When is inventory the cheaper insurance than qualifying a supplier, and what does it fail to protect against? (The harder one: weigh how long the outage lasts against how long a buffer lasts, and what a permanent loss of a supplier does that a buffer cannot cover.)
A supplier offers to hold safety stock for you and promise priority allocation in a shortage, for a fee, so you can stay single-source. When does that “assured supply” substitute for a second source, and what is it still exposed to?

10. Ship it — your decision memo

Write a one-page memo to Kestrel’s leadership. State the call (begin qualifying a regional second source for the vision-sensor module and decline the exclusive single-source lock-in — while not dual-sourcing the low-impact commodity parts). Show the reasoning in two or three lines (a strategic, high-impact part whose exposure just rose with the tariff and single-region concentration; the expected cost of a line-down dominates the certain discount; a second source earns its hidden costs here because impact and exposure are high). Name what you rejected — taking the discount and deepening the single point of failure, and, on the other side, dual-sourcing everything regardless of impact. Name the one thing that would change your mind (a much lower disruption probability, a cheaper way to cover the impact such as bonded buffer stock, or a second source whose risk is correlated with the first). Keep it to a single page. This memo is your own argued claim — not a credential.

11. Sources

Meridian Devices and Kestrel Robotics, and every figure attached to them — volumes, prices, outage weeks, and disruption probabilities — are composite and illustrative, constructed for clean teaching reasoning, not drawn from or claimed about any real company. The frameworks used to reason about them are standard supply-chain and procurement concepts; references below.

Concept / claim	Source (publisher)	URL	Accessed
Kraljic purchasing portfolio (profit impact × supply risk quadrants)	Wikipedia — Kraljic matrix	https://en.wikipedia.org/wiki/Kraljic_matrix	2026-07-19
Second-sourcing a critical component in electronics	Wikipedia — Second source	https://en.wikipedia.org/wiki/Second_source	2026-07-19
Supply-chain risk management, disruption exposure	Wikipedia — Supply chain risk management	https://en.wikipedia.org/wiki/Supply_chain_risk_management	2026-07-19
Multi-sourcing across more than one supplier	Wikipedia — Multisourcing	https://en.wikipedia.org/wiki/Multisourcing	2026-07-19
Single point of failure (one supplier halts the flow)	Wikipedia — Single point of failure	https://en.wikipedia.org/wiki/Single_point_of_failure	2026-07-19
Risk as probability × impact (risk matrix)	Wikipedia — Risk matrix	https://en.wikipedia.org/wiki/Risk_matrix	2026-07-19
Expected value (probability-weighted cost)	Wikipedia — Expected value	https://en.wikipedia.org/wiki/Expected_value	2026-07-19
Total cost of ownership beyond unit price	Wikipedia — Total cost of ownership	https://en.wikipedia.org/wiki/Total_cost_of_ownership	2026-07-19
Continuity planning against supplier disruption	Wikipedia — Business continuity planning	https://en.wikipedia.org/wiki/Business_continuity_planning	2026-07-19

Next up

That’s the top of the Supply Chain & Procurement track. Browse all courses → to pick your next call.