There is a fork in every aging website, and the money is decided by which branch you take. On one side, a refresh: same bones, new skin, live in weeks. On the other, a rebuild: new structure from the ground up, live in months, some of it rebuilt worse before it is rebuilt better. Owners get pushed toward rebuilds more than they should — rebuilds are bigger invoices and easier to sell. So here is the honest test, written down the way we use it ourselves.
The three-question test
Does the site's structure still match the business? If the pages answer the questions customers actually have, the structure is fine and the problem is above the foundation. If the site cannot represent a service you now sell, or a location you now have, the structure is the problem — and structure is what a refresh cannot fix.
Does anything under the hood still work as promised? Forms arriving, pages loading fast, the address resolving, search engines reading it. If the mechanics are sound, the mechanics should stay. If the last vendor left and the parts are held together with tape, a reskin would be putting new paint on a car that needs its engine opened — inspect first, then decide.
What does the business need the site to do next year? If the answer is "the same, but better looking," you are in refresh territory. If the answer introduces booking, payments, a catalogue, or three audiences where there was one, you are describing a site your current build was never made to hold.
What a refresh honestly is
In our hands: the design system comes up to today — typography, spacing, colour, contrast, phone layouts — while the addresses, the pages that rank, and the words that work stay exactly where they are. The skin changes; the equity does not move. Every URL anybody has ever shared keeps working, which matters more than owners are usually told: the fastest way to break a business's search presence is to come out of a redesign with a new map of addresses and no redirects.
A refresh is also honest about its ceiling. It will not fix a story problem, a structure problem, or a "what is this business actually selling" problem. It fixes tired. Tired is common, and it is worth fixing.
What a rebuild is for
A rebuild earns its cost when the site is asked to do something new — carry multiple locations, sell, take bookings, run follow-up, speak to more than one audience. It is also the right call when the platform itself is the trap: a rented builder that cannot be left without abandoning everything, or a stack nobody can service. In those cases a reskin would be paying to keep a cage tidy.
Rebuilds have one more honest property: they hurt mid-flight. The professional version tries to keep the old site alive underneath while the new one is built — which is why a real rebuild is measured in weeks with a date you can plan around, not a weekend with a surprise in it.
The quiet middle
Most aging sites that reach us need neither the full ceremony of a rebuild nor another bandage. They need the middle path: redo the vanishing sections first — the hero, the enquiry path, the phone layouts — then page by page as the business can absorb. The site never goes dark, nothing anybody shared stops working, and every change is judged by the one test that matters: did more of the right people reach out.
If you take one thing from this: ask any vendor which branch they are recommending and why, and listen for whether they can describe the case against their own answer. A rebuild pitch that cannot say when a refresh would have been right is not advice. It is a quote looking for a project.