When to redesign

The triggers that justify it

A site that does not work on a phone. That is the clearest case, because it fails for most of the people trying to use it and no amount of editing content fixes a structural problem of that kind.

A platform being retired or no longer supported, which is a deadline rather than a choice. Or a business that has genuinely changed: a new trade, a new market, or a service list that no longer resembles what the site describes.

And a site nobody can edit. If a change requires finding a developer who has moved on, every small correction is deferred forever, and the site decays because updating it is too much trouble.

The reasons that look convincing

Age. A site built five years ago that works on a phone, loads quickly and describes your business accurately is not failing at anything, and replacing it is a substantial cost for an aesthetic preference.

Boredom. You look at your own site far more than any customer ever will, so what reads as tired to you is entirely new to somebody arriving for the first time this afternoon.

A competitor's new site. Their rebuild tells you about their budget and nothing about whether yours is working, and the honest comparison is your enquiry count against your own last year.

Diagnose before you decide

Most complaints about a website are content problems wearing a design costume. Unclear wording, a missing price, no obvious way to make contact, and a homepage that does not say what the business does.

All four are fixable in an afternoon on the site you already have. Our page on redesign or refresh covers the distinction and what each one actually involves.

So write down the specific thing that is wrong before deciding what to do about it. A rebuild justified by a vague sense that the site is dated tends to reproduce the same faults in newer styling.

What actually justifies a rebuild

The list below is the complete set of genuine triggers. If none applies, the honest answer is that a refresh will get you most of the benefit for a fraction of the cost and risk.

• It does not work properly on a phone
• The platform is retired or unsupported
• The business has substantially changed
• Nobody available can edit it
• It is slow in a way content cannot fix
• It excludes people who cannot use a mouse
• The structure cannot hold what you now offer

What a rebuild puts at risk

Search positions earned over years, if page addresses change without redirects. That is the single most expensive mistake in a website rebuild and it is entirely avoidable with an afternoon of planning.

Our page on redirects covers the mechanics. Every old address needs to point somewhere sensible, and a map of them should exist before anything is switched over.

Content is also at risk. Pages that were quietly bringing enquiries get dropped in a rebuild because nobody checked which ones were working before deciding what the new site would contain.

Timing it sensibly

Not during your busiest season. A rebuild takes attention from the business and something always goes wrong in the first week, which is the worst possible time for that week to be a trading peak.

Not immediately before a campaign either. Launch the site, let it settle for a few weeks, confirm the contact routes work and the numbers look normal, and only then start promoting it.

Our page on the launch checklist covers that settling period. The gap between launch and promotion is where avoidable problems get found.

Rebuild what needs it, keep what works

A rebuild does not have to be total. New structure and design with the existing content carried across is frequently the right scope, and it protects the pages that were already earning their place.

Take the measurements first. Knowing which pages bring arrivals and enquiries turns a rebuild from a fresh start into an improvement on something, which is a genuinely useful difference.

And keep the old site accessible somewhere until the new one has proved itself. Being able to check what a page used to say settles a surprising number of questions in the first month.