The websites and applications that become expensive to own rarely fail technically. They become expensive because the original decision optimised for launching quickly, and nobody asked what maintaining it would look like once the person who built it had moved on. A few questions at the start prevent most of it.

Who maintains this in two years?

If the answer is a single specialist, an agency, or a framework few people use, you have accepted a constraint that will eventually cost you. Widely-used, unremarkable technology has a large pool of people who can work on it and abundant documentation. That matters more than elegance for anything you intend to keep.

What happens if the vendor changes?

Managed platforms and hosted services are frequently the right choice, but check the exit before committing. Can you export your content and data in a usable format? Does the code run elsewhere? Are you able to move your domain and email independently? A platform that holds your data in a proprietary format with no meaningful export is a decision that is difficult to reverse later.

Own the things that are genuinely yours

  • Your domain, registered in the company's name with access you control.
  • Your DNS, and the ability to change it without asking anyone.
  • Your source code, in a repository you own.
  • Your data, with exports you have actually tested.
  • Your analytics and customer communication channels.

Match complexity to the actual problem

A brochure site does not need the architecture of an application. An application with a hundred users does not need the architecture of one with a million. Complexity added for hypothetical future scale is paid for immediately, in build time and maintenance, and the anticipated scale frequently never arrives. Build for the problem you have, and leave a clean path to the one you might.

Cost that appears later

Look past the monthly figure. Ask what happens as traffic grows, what the egress and storage charges look like, what falls outside the plan, and what the renewal looks like. The cheapest option at launch is regularly not the cheapest at year three, and the difference is usually discovered at the worst moment.

Write down the reasoning

A short document recording what you chose and why is disproportionately valuable. In two years, when someone asks why this database or that platform, the answer should not depend on whether a particular person still works there.