The build-vs-buy debate is usually a vibes argument. The engineers want to build because they think they can do it better, faster, and they enjoy the work. The CFO wants to buy because the line item is predictable. The CEO wants whichever option lets them ship the thing the customer is asking about. Without a framework, the debate goes in circles and the decision gets made by whoever spoke last.
Three questions, in order. First: is this our core differentiator, the thing customers come to us for? If yes, build it. Owning the differentiator end-to-end is what makes you durable. If no, the answer is probably buy. Most companies have one or two real differentiators. Everything else is plumbing. Buy your plumbing from companies that specialize in plumbing.
Second: does an acceptable off-the-shelf option exist today? "Acceptable" means it does 80% of what you need, integrates with your existing stack, and is run by a company that will still be around in three years. If yes, the buy case strengthens — the time-to-value is weeks, not quarters, and the maintenance burden lives with the vendor. If no acceptable option exists, you may have to build, but proceed with the third question.
Third: are you sure no acceptable option exists, or are you about to build a slightly different version of something that already exists in 30 SaaS tools you haven't looked at? The number of times we've watched a team spend six months building an internal tool that does roughly what Notion + a Zapier integration would have done — at one-tenth the maintenance cost — is much higher than you'd think. The build case has to clear that bar specifically: not just "we could build this," but "we have looked at the obvious off-the-shelf options and they genuinely won't work."