When not to rebuild your website
We're asked for rebuilds constantly. A fair share of the time, the honest recommendation is a much smaller piece of work.
A rebuild is a large, slow, expensive thing. It is also the first solution most people reach for when a site is underperforming, because it feels decisive.
Before agreeing to one, we run an audit. Often it comes back saying the structure is fine and something much smaller is doing the damage.
Things that masquerade as needing a rebuild
- Hosting in the wrong region, adding half a second to every request.
- Unresized images doing most of the page weight on their own.
- A dozen marketing scripts loading before the content.
- A homepage that never says what the company sells.
- A contact form that has been quietly failing for months.
That last one is worth checking today. We have found silently broken forms on sites whose owners assumed business was simply slow.
The cheapest project is the one that fixes the actual problem, and it is rarely the one that was requested.
When a rebuild genuinely is the answer
When the platform can no longer do what the business needs, when nobody can safely deploy a change, or when the underlying dependencies are old enough to be a security problem. Those are real reasons, and in those cases we say so and get on with it.
- #audit
- #performance
- #advice
Keep reading
Build for the phone people have, not the one on your desk
Most performance advice assumes a fast device on a fast network. Here is what changes when you design for a mid-range Android on patchy data.
ReadWorking togetherWhy software quotes double, and how to stop yours doing it
An honest account of what drives the cost of a build, written from the side of the table that produces the estimates.
Read