The core idea
Markets coordinate by price; firms coordinate by authority. Each has a cost. You should make rather than buy when transacting in the open market would be more expensive than running the activity yourself — typically because the activity demands relationship-specific investments, repeats often under uncertainty, and would expose you to hold-up by a thin set of counterparties. When those conditions are absent, the market is faster, cheaper and more disciplined than your own org chart. — after Coase (1937) and Williamson (1985)
The hero diagram
Where to draw the firm's boundary
Walk a candidate activity down the tree. The price mechanism is the default — only step inside the firm when the transaction itself would misbehave on the open market.
Is the activity strategically core — a source of differentiation, learning or rents?
-
No — it is generic, commoditised inputBuy. Use the market; let suppliers compete on price and innovation.
-
Yes — it shapes how you compete
Does it require relationship-specific assets — dedicated capacity, custom tooling, co-located plant, proprietary know-how?
-
Low specificity — assets are redeployableBuy with a standard contract. Switching is cheap; the market disciplines the supplier.
-
High specificity — once built, the asset only has value in this relationship
Is the supplier market thin and the future hard to specify — few credible counterparties, high uncertainty, frequent re-negotiation?
-
Thick market, low uncertaintyBuy with safeguards. Long-term contract, dual sourcing, or reciprocal hostages.
-
Thin market, high uncertainty — hold-up risk is real
Can you absorb the activity without losing the market's scale, focus and incentive intensity?
-
No — internalising would dull incentives or sub-scale the activityAlly. Joint venture, equity stake, or deep partnership — share the specific assets, share the risk.
-
Yes — you can run it at scale and keep it sharpMake. Bring it inside; coordinate by authority rather than by price.
-
-
-
Frameworks in this module
Named ideas to remember.
How to apply
For any activity you are considering internalising or outsourcing.
- Ask: is this core to how we compete? Generic inputs belong in the market. Only activities that shape your competitive position warrant the make analysis.
- Measure asset specificity. If the supplier builds something only valuable in this relationship, hold-up risk is real.
- Count credible counterparties. A thin market with one or two viable suppliers dramatically raises transaction costs.
- Estimate the cost of bringing it inside. Will internalisation dull incentives, sub-scale the activity, or distract management? The firm is not free.
- Consider the alliance option. When both make and buy are costly, a joint venture or deep partnership can share the specific investment without full integration.
Key reading · Coase · Economica 1937
The Nature of the Firm.
Coase's 1937 paper asks a deceptively simple question: if markets are efficient, why do firms exist at all? His answer is that using the price mechanism itself has costs — discovering prices, negotiating contracts, bearing uncertainty. When those costs exceed the cost of internal coordination, the firm expands its boundary. This insight remains the foundation of every make-or-buy analysis.
Markets are not free. Neither are firms. The boundary goes where the costs cross.