Key takeaway
Treat build-vs-buy as a test of whether the capability is core to your differentiation, not whether your engineers can build it. The post recommends buying by default and building only when the capability is a durable differentiator, no vendor adequately fits, and you can carry the maintenance forever.
"Can we build it?" is almost always yes — good engineers can build nearly anything, which is exactly why it is the wrong question. The right one is whether this capability is core enough to your differentiation to justify owning it forever, when a vendor would do it now. Here is the worked version: the same seven-section structure YourBrief generates, applied to build-vs-buy — with a decision matrix you can run against your own case.
| Factor | Lean BUILD when… | Lean BUY when… |
|---|---|---|
| Differentiation | it is a reason customers choose you | it is table-stakes plumbing everyone needs |
| Vendor fit | no vendor fits, or all are genuinely bad | a solid vendor exists at a fair price |
| 3-year total cost | owning it beats the vendor over that window | subscription beats fully-loaded build + upkeep |
| Speed | you can wait quarters for it | you need it this quarter |
| Lock-in | you must control the roadmap yourself | integration is standard and reversible |
The decision
Building is close to a one-way door: you own the code, the maintenance, the on-call, and the hiring around it, effectively forever. Buying stays more reversible — as long as you keep the integration standard and avoid deep lock-in. So the decision is not "can we build it?" Your team probably can. It is "is this capability core enough to our differentiation to justify owning it forever, when a vendor would deliver it now and free the team for the roadmap?"
Key questions to answer before deciding
- Is this capability a reason customers choose you (core), or table-stakes everyone needs (context)?
- Does a real vendor solution actually fit, or have you only compared against a strawman?
- What is the fully-loaded three-year cost of building — including maintenance, on-call, and the roadmap work those engineers will not do?
- How fast do you actually need it, and can the team afford the build without starving the core product?
- What is the lock-in and switching cost of the buy option?
Recommended frameworks
Core versus context (Geoffrey Moore). Build the core — the capability that differentiates you — and buy the context, the necessary plumbing that does not. Most "we should build this" arguments are really context wearing a core costume.
Three-year total cost of ownership. Compare the vendor subscription against the fully-loaded cost of building AND maintaining: initial build, plus the ongoing engineers, on-call, and opportunity cost of the roadmap they are not shipping. The initial build is the visible cost; the maintenance tail is the one teams tend to underweight.
Reversibility and lock-in check. Score the buy option on how cleanly you could leave it. A vendor you can swap behind a standard interface is a two-way door; one whose data and workflow you cannot extract is closer to a one-way door — price that risk in.
Decision criteria
Build only when the capability is a durable differentiator, no vendor adequately fits, and you can carry the maintenance and on-call forever. If it is context rather than core, buy it even if building would be more interesting. Otherwise the default is buy.
Sources to consult
Your own roadmap (name exactly what building this displaces); a genuine vendor evaluation, not a strawman built to lose; one team that built the thing you are considering and what maintaining it actually cost them a year later.
Next steps
Classify the capability as core or context; run a real vendor evaluation; compute the three-year TCO including the maintenance tail and opportunity cost; default to buy unless the capability clears the durable-differentiator bar.
When to escalate
Escalate to engineering leadership or the board when building would require standing up a new team or competency, when the buy option creates deep lock-in to a vendor critical to your operation, or when the capability is strategic enough that owning it is genuinely a board-level bet.
The honest answer is often "buy it and move on" — the cases genuinely worth building are fewer than they first appear. Generate this exact brief against your own capability, vendor options, and TCO — $1 to start.