Every few months a new framework, database, or deployment paradigm shows up on Hacker News with momentum behind it. The benchmarks look good. The syntax is elegant. The community is enthusiastic. And somewhere, a builder with a half-finished product thinks: maybe I should switch.
Don't.
The argument for boring technology is not an argument against progress. It's an argument about timing. Early-stage products fail for dozens of reasons — picking the wrong audience, building the wrong feature, running out of money. Almost none of them fail because the builder chose a database that was ten years old instead of ten months old.
When something breaks at 2 a.m. — and it will — you need two things: a clear error message and a Stack Overflow thread with an accepted answer from 2019. Boring technology gives you both.
Newer tools carry a debugging tax. The error messages are less documented. The edge cases are less explored. The community answers, when they exist, start with "I think…" instead of "Here's exactly what happened."
Every hour you spend deciphering an unfamiliar stack trace is an hour you didn't spend talking to a customer, shipping a fix, or sleeping. For a team of one or two, that tax compounds fast. The surface area of things that can go wrong should be as small as you can make it — not because you're afraid of complexity, but because your attention is finite and your customers don't care what's under the hood.
If your product survives long enough to need a second engineer, you want someone who can contribute on day one, not day thirty. The more conventional your stack, the shorter the ramp.
This matters more than most builders admit. A team of two has no slack for onboarding. There's no internal wiki worth reading. There's no senior engineer to pair with. The new person needs to open the codebase, recognize the patterns, and start fixing bugs.
Exotic technology choices turn hiring into a filter that works against you. You're not selecting for quality — you're selecting for familiarity with a niche tool. That narrows the pool without improving the outcome.
Here's the strongest case for boring: it lets you ship features instead of fighting infrastructure.
Think about the last three things your customers asked for. A better notification, a cleaner export, a way to invite teammates. None of those required a novel runtime or a distributed consensus algorithm. They required your time and attention directed at the problem instead of the plumbing.
Boring technology disappears. It becomes transparent. You stop thinking about it and start thinking about what the customer actually needs. That shift — from infrastructure focus to customer focus — is the single highest-value move an early-stage builder can make.
When the infrastructure is predictable, the surprises come from the market, where they belong.
Before adopting any technology, run it through one question: Does this choice directly improve a customer outcome?
If yes — the customer gets faster responses, more reliable data, lower latency they can feel — then the newer tool might be worth the cost. Sometimes boring isn't good enough. Real-time collaboration has different demands than a settings page. Acknowledge that.
But if the answer is "it improves developer experience" or "it's more elegant" or "it's what I want to learn," those are honest reasons but they are not customer reasons. Default to the option with the longest track record. The one with the most battle scars. The one nobody writes blog posts about anymore because it just works.
There's a quieter force at play. Builders often tie their identity to their tools. Choosing something old can feel like admitting you're behind. It can feel like a lack of ambition.
Flip that framing. Choosing boring technology is a statement about what you value: customer outcomes over technical novelty. Shipping over exploring. Results over résumé items.
The best products are not interesting to operate. They are interesting to use. The technology underneath should be so unremarkable that the only thing worth talking about is what the product does for the person on the other side of the screen.
Pick the tool that lets you forget about infrastructure and remember your customer. Pick the tool your next hire already knows. Pick the tool whose failure modes are documented in plain language by people who hit them five years ago.
New technology is a bet. Boring technology is a known quantity. Early-stage builders don't need more bets — they need fewer, better ones, placed where they actually matter: on the product, and on the customer.
Be the first to comment.
0 comments
Loading comments...