Founders lose weeks agonizing over build-vs-buy decisions that senior operators settle in one meeting. Not because the seniors are smarter — because they're using a shorter checklist. Here it is.
The four questions
1. Is this your product, or is it plumbing? If customers pay you for this thing specifically, lean build — it's where your differentiation lives. If it's auth, billing, email, analytics, search, or notifications: it's plumbing. Buy plumbing. Nobody ever chose a product because it had artisanal password reset.
2. What does your time cost? The build estimate is never the real cost. The real cost is build + maintain + secure + document + the features you didn't ship while doing all that. Multiply the estimate by three, then ask whether the vendor's price still looks expensive.
3. Can you leave? The real risk of buying isn't the subscription fee — it's lock-in. Before adopting a vendor, ask: if this company dies, doubles its price, or turns hostile, what does migration cost? Data export, standard protocols, and open-source options all lower the exit toll. A vendor you can leave is safe to marry.
4. Is the requirement actually standard? Buying wins when your need matches what the tool does out of the box. If you're on page four of the "customization" docs before covering your core case, you're not buying a product — you're building on someone else's foundation. That's the worst of both worlds.
Hidden costs people miss
On the build side: on-call burden, security patching forever, the bus factor when the engineer who wrote it leaves, and the quiet way internal tools rot without a paying customer to demand quality.
On the buy side: per-seat pricing that scales painfully with your team, integration work that's always more than the demo suggested, and roadmap hostage situations — the feature you need is "coming next quarter" for six consecutive quarters.
The startup default
Buy everything except the thing you sell. Pre–product-market fit, engineering hours are the scarcest resource in the company. Every hour on plumbing is an hour not spent finding out whether anyone wants your product. You can revisit every one of these decisions later, from a position of scale and knowledge. The startups that die aren't the ones that overpaid for Stripe.
Two exceptions worth knowing
- When the vendor bill threatens the business model. If COGS from a vendor makes your margins impossible at scale, building becomes a strategic project — schedule it deliberately, after PMF, not in a panic.
- When "buy" means "enterprise sales cycle." If procurement plus integration exceeds the honest build cost for something simple, build the simple version. Emphasis on simple — and honest.
The 20-minute meeting
Put the four questions on a whiteboard. Answer them with the most senior technical judgment in the room. Write the decision down with its reasoning — future-you will want to know why, because the answer can change as you scale, and that's fine. What's not fine is relitigating it every quarter from scratch.