SOP: Buy a tool vs build it in-house

SOP: Buy a tool vs build it in-house

Run this whenever someone proposes building internal tooling.

Stage 1 — Define the need precisely (owner: ops lead)
☐ Is this core to your edge, or undifferentiated plumbing?
☐ How many off-the-shelf tools already do 80% of it?
☐ What's the true cost of maintaining a build (not just building it)?

Stage 2 — Apply the rule
☐ Buy: undifferentiated, well-served by vendors, and you'd rather spend the team's time elsewhere. Cost: subscription, vendor lock-in, missing the last 20%.
☐ Build: it's a real differentiator, or no tool fits and the gap hurts. Cost: ongoing maintenance forever, and the original builder leaving.

Stage 3 — Guardrails (owner: lead)
☐ Buy → check export/data portability before signing.
☐ Build → name the long-term maintainer, not just the builder.
☐ Default to buy for plumbing; reserve build for your moat.
☐ Done-when: total cost of ownership compared, not just upfront price.

Trigger: "let's just build it." Done-when: decision and owner recorded with TCO.

Save this. Run it every time.