Make it or buy it: the unit-economics call, and the volume that flips it

1. Before you start

A make-or-buy call compares what a part costs to make in-house against what it costs to buy from a supplier — and the only costs that belong in it are the ones that actually change when you decide. A tiny example: suppose making a bracket costs $6 of metal and labour you would spend only if you make it, plus a $12,000-a-year machine you would rent only if you make it; buying the same bracket costs $8. Whether making is cheaper depends entirely on how many you need — the $12,000 is one lump, and it spreads thinner the more brackets you make.

Three honest statements before you start:

This is a Decide course. You read a situation, learn the idea, and make a call. You do not write or run any code.
The companies below, Cascade Actuation and Northwind Cycles, are composite — invented firms built from ordinary figures so the arithmetic is clean. No number is a claim about any real company.
This is not a certification. It shows, to you, that you can defend a make-or-buy call.

You need only arithmetic. The hard part is deciding which costs count.

2. The Situation

Cascade Actuation builds industrial linear actuators and buys the machined gear housing inside each one from an outside shop at $50 a unit. Operations wants to bring it in-house: “our own cost sheet says we can make it for less, so we are handing the supplier free margin.” Finance pushes back: “once you load in the plant overhead, making it costs more than $50 — stay out.”

They cannot both be right, and the number that settles it is one neither of them named: how many housings Cascade needs a year. Get the make-or-buy call wrong and you either sink a dedicated machine cell that never pays for itself, or you hand a supplier margin you could have kept.

3. What you’ll be able to do

After this course you will be able to:

Build the relevant cost to make a part — its variable cost plus only the fixed cost the make decision actually adds — and compare it to the buy price.
Spot the two symmetric traps: comparing the buy price to variable cost alone (which flatters making) and to fully-loaded cost (which double-counts sunk overhead and flatters buying).
Find the volume break-even where the make-or-buy call flips, and name the number that moves it.
Layer the strategic caveats — capability, IP, supplier risk — on top of the number as judgment, without letting them replace the arithmetic.

4. Prerequisites & time box

Prerequisites: arithmetic, and the idea of fixed vs variable cost (defined in section 6). No spreadsheet, no accounting background, no code setup — the Decide hall is read-and-decide in the browser.

Time box: about 19 minutes of reading (measured), plus your own thinking time on the call. That is under the 25-minute cap for a concept course.

Difficulty: 5 / 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 signup, no paid tool, requires_gpu: false.

5. The case & where the numbers come from

Cascade Actuation is a composite manufacturer: a mid-size actuator maker whose cost structure is built from ordinary figures a machine shop would recognise, chosen for clean arithmetic and not drawn from or claimed about any real firm. The definitions — fixed cost, variable cost, sunk cost, relevant cost, break-even — are standard and cited in section 11. Every dollar below is an in-course assumption; every later number is computed from these.

Item	Figure
Buy price per housing (current supplier)	$50
Make: direct material per housing	$20
Make: direct labour per housing	$12
Make: variable overhead per housing	$3
Dedicated fixed cost to make (CNC cell lease + setup technician), per year	$180,000
Allocated existing plant overhead on the cost sheet, per housing	$10
Annual requirement	15,000 housings

The dedicated fixed cost is the cost of the machine cell and the technician Cascade would add only if it makes the housing — it appears only under the make option. The allocated plant overhead is a slice of costs Cascade already carries (the plant manager, the building, shared depreciation) that its accounting spreads across every part; it is on the housing’s cost sheet whether or not Cascade makes it.

6. The Concepts

Relevant cost to make

The cost of making a housing has two parts. The first is its variable cost — the material, labour, and variable overhead each housing causes: $20 + $12 + $3 = $35 a unit. The second is the dedicated fixed cost the make decision adds: the $180,000-a-year CNC cell and technician that exist only if Cascade makes the part. Spread over the 15,000 housings Cascade needs, that is $180,000 ÷ 15,000 = $12 a unit.

So the relevant cost to make — the cost that actually changes if Cascade makes rather than buys — is $35 + $12 = $47 a housing, against the $50 buy price. On the numbers, making is $3 a unit cheaper: 15,000 × $3 = $45,000 a year kept in-house. The two costs that change are the variable cost and the dedicated fixed cost; every other cost on the sheet is the same whether Cascade makes or buys, so it does not belong in the comparison.

The variable-cost-only trap

Operations looked at the $35 variable cost, saw it sitting far below the $50 buy price, and concluded Cascade saves $15 a unit by making — “the supplier is charging us $50 for something that costs us $35.” That ignores the $180,000 dedicated cell the make option adds. At 15,000 units the cell adds $12 a unit, so the real saving is $3, not $15. Worse, the $180,000 is a lump: at a smaller volume it spreads over fewer units and swamps the $15 gap. Comparing the buy price to variable cost alone always makes in-sourcing look cheaper than it is, because it leaves the dedicated fixed cost out entirely.

The fully-loaded-cost trap

Finance made the opposite error. Its cost sheet loads all of a housing’s costs together: $35 variable + $12 dedicated cell + $10 allocated plant overhead = $57 a unit — above the $50 buy price, “so making loses us $7 a unit.” But that $10 is an allocation of costs Cascade pays regardless — the plant manager’s salary, the building, the shared machines — and outsourcing the housing does not make any of them go away. That $10 is a sunk cost to this decision: it is incurred whether Cascade makes or buys, so it must not swing the call. Strip it out and the relevant make cost is back to $47, below the $50 buy price. Fully-loaded cost is the right lens for pricing the whole product line over the long run; it is the wrong lens for a make-or-buy call, because it double-counts overhead that the decision cannot change.

The volume break-even

Both traps disappear once you ask the question the argument was really about: at what volume do making and buying cost the same? Making beats buying only when the $180,000 dedicated cell is spread over enough units to more than cover the $15 gap between the $50 buy price and the $35 variable cost. That volume is the break-even:

Dedicated fixed cost ÷ (buy price − variable cost) = $180,000 ÷ ($50 − $35) = $180,000 ÷ $15 = 12,000 housings a year.

Below 12,000, buy; above 12,000, make. Cascade needs 15,000, which is past the break-even, so making wins — by the $45,000 computed above. But watch how thin the margin is: if the annual requirement were 10,000 instead, making would cost 10,000 × $35 + $180,000 = $530,000 ($53 a unit) while buying costs 10,000 × $50 = $500,000 — and buying wins by $30,000. Same part, same cost sheet, opposite call. The fact that flips it is the annual volume.

(An interactive calculator sits here — enter the buy price, the three make cost components, the dedicated fixed cost, the allocated overhead, and the annual volume, and it returns the relevant make cost, the fully-loaded cost, the break-even volume, and the yearly make-versus-buy difference, so you can watch the call flip.)

Strategic caveats: capability, IP, supplier risk

The number sets the floor of the argument, not the whole of it. Three caveats sit on top of it as judgment. Capability: making the part means building and keeping a skill the plant may not have, and losing it if the cell is later shut. IP: if the part carries know-how you would rather a competitor never see, outsourcing hands it to a supplier — a cost the dollar figure does not show. Supplier risk: buying ties you to another firm’s capacity, quality, and lead time, and a single-source supplier can hold you up. These can override a small cost gap — you might in-source a part the numbers say to buy because it is core IP, or dual-source a critical part despite the premium. What they must not do is replace the arithmetic: you decide by naming the measured cost gap first, then asking whether the strategic reason is worth that specific number.

7. Your Call

You have seen how relevant cost, the two traps, and the break-even volume settle Cascade’s housing call. Now a different one lands on your desk.

Northwind Cycles assembles e-bikes and buys its battery packs from a supplier at $220 a pack. Its own build estimate is material $150 + labour $30 + variable overhead $10 = $190 a pack, and a dedicated assembly line with a test rig and technician would add $420,000 a year. The cost sheet also allocates $15 a pack of existing plant overhead. Marketing forecasts 11,000 packs a year. The pack also holds Northwind’s own cell-balancing know-how, and only one supplier can currently build it.

How this differs from the taught case (the transfer): this is a different composite company and sector — e-bike assembly, not Cascade’s actuator machining — the figures are different so the arithmetic must be redone, and it adds a new constraint: the strategic layer (core IP plus a single-source supplier) now sits against a cost gap that points the other way, and you must decide with it, not around it. The core concept is the same: relevant make cost versus buy price, and the break-even volume that flips the call.

8. Self-check

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

Why is “our cost sheet says $57 fully-loaded, so buying at $50 is cheaper” the wrong test for a make-or-buy call?
Which two costs actually change when you make instead of buy, and which one on the cost sheet does not?
How do you turn the dedicated fixed cost and the per-unit gap into the volume at which the call flips?

If any is fuzzy, reread section 6 — relevant cost, the two traps, and the break-even volume are the whole course.

9. Stretch

Push the decision further on your own:

Back at Cascade: at what buy price would making and buying tie at the 15,000-unit volume, given the $35 variable cost and the $180,000 cell? (Set 15,000 × ($P − $35) against $180,000 the other way, or just read it off the relevant make cost.)
The genuinely harder one: suppose half the $180,000 dedicated cell is a lease Cascade can cancel each year and half is a machine it has already bought and cannot resell. Which half belongs in next year’s make-or-buy call, and how does that change the break-even volume?
Write the one sentence you would say to an operations lead who insists “our cost sheet proves we should make it.”

10. Ship it — your decision memo

Write a one-page memo to Northwind’s operations director. State the call (at 11,000 packs the relevant make cost is about $228 against a $220 buy price, so on cost alone, buy — while naming the strategic reason to reconsider). Show the two-line arithmetic (relevant make cost = $190

$420,000 ÷ 11,000; break-even = $420,000 ÷ $30 = 14,000 packs, above the 11,000 forecast). Name what you rejected (the variable-cost-only “we save $30 a pack” claim; the fully-loaded-cost lens that counts sunk overhead) and why. Name the one thing that would change your mind (volume rising past 14,000 packs, or the IP and single-supplier risk outweighing the roughly $90,000 gap). Keep it to a single page a director grasps in two minutes. This memo is your own argued claim — not a credential.

11. Sources

Cascade Actuation and Northwind Cycles, and every dollar figure attached to them, are composite and illustrative — constructed for clean teaching arithmetic, not drawn from or claimed about any real company. The concept definitions used to reason about them are standard; references below.

Concept / claim	Source (publisher)	URL	Accessed
Make-or-buy is decided on the costs that change (outsourcing decision)	Wikipedia — Outsourcing	https://en.wikipedia.org/wiki/Outsourcing	2026-07-20
Relevant cost = the cost that changes with the decision	Wikipedia — Relevant cost	https://en.wikipedia.org/wiki/Relevant_cost	2026-07-20
Sunk / allocated cost does not belong in the decision	Wikipedia — Sunk cost	https://en.wikipedia.org/wiki/Sunk_cost	2026-07-20
Definition of a fixed cost	Wikipedia — Fixed cost	https://en.wikipedia.org/wiki/Fixed_cost	2026-07-20
Definition of a variable cost	Wikipedia — Variable cost	https://en.wikipedia.org/wiki/Variable_cost	2026-07-20
Break-even volume = fixed cost ÷ per-unit margin	Wikipedia — Break-even point	https://en.wikipedia.org/wiki/Break-even_point	2026-07-20
In-house vs supplier as a vertical-integration choice	Wikipedia — Vertical integration	https://en.wikipedia.org/wiki/Vertical_integration	2026-07-20

Next up

Finished this call? Continue the Operations & Process track:

Running hot: why an 80%-full clinic waits, and a 95%-full one drowns → · Browse all courses