A sourcing lead lines up the make-or-buy numbers: in-house conversion comes in 8% higher than the supplier's quote. The spreadsheet says buy. Eighteen months later, the product needs a minor module change, and the team can't make it without re-qualifying the supplier, reopening the contract, and waiting two quarters. The 8% they saved is now the tax on every change they want to make.
Cost is the easy half
In modular sourcing, make-or-buy is usually presented as a cost model: build it or buy it, pick the cheaper one. That comparison is real, but it measures one moment. What it misses is the second-order effect — what you choose to make is what you keep the ability to change.
A module you build in-house sits inside your own process. You can revise it, swap a component, respond to a design change on your own clock. A module you buy sits inside the supplier's process, governed by their lead time and their willingness to touch it.
The decision is about control, not price
The useful question is not "which is cheaper." It is "which changes do I need to own." Three signals push a module toward make:
- The module sits on the critical path of frequent product updates.
- The specification is still moving and will keep moving.
- The capability is core — losing it would hollow out the team's ability to design the next generation.
Three signals push toward buy:
- The module is stable, standardized, and far from the product's differentiating edge.
- Several capable suppliers exist and switching is real.
- Holding the capability in-house would distract from where the team actually adds value.
Make-or-buy is less a purchase and more a statement about which capabilities the company intends to keep. A wrong call doesn't show up in the cost model. It shows up in the third revision no one can deliver.
The modular trap
Modular sourcing makes make-or-buy feel tidy: each module is a clean building block, so the logic goes, just source each one where it is cheapest. But modularity cuts both ways. The more you buy, the more your product architecture depends on suppliers you don't control — and the harder a cross-module change becomes, because it now needs agreement from several independent parties.
A buyer we know kept make-or-buy purely financial and outsourced a module that later became central to a new variant. Reclaiming it took a year of capability rebuild. The original saving was real. The recovery cost was larger, and invisible to the model that approved the buy.
What good looks like
Mature teams run make-or-buy as a periodic review, not a one-time call at launch. They score each module on stability and strategic weight, and they revisit the answer when the product enters a new phase. A module that was right to buy at launch can be right to bring in-house two years later, once volume and clarity arrive.
The line to remember
What you make, you can change. What you buy, you rent — including the ability to adapt. Make-or-buy in modular sourcing is a capability decision dressed up as a cost call; treat it as the former and the architecture stays yours.