The default answer is no
Most sites brought to us for redesign do not need one. They need one part fixed: the message, the mobile experience, the three templates carrying commercial traffic, or the technical foundation that makes everything else invisible.
That matters financially because a targeted fix costs a fraction of a rebuild and preserves everything already working. A redesign resets accumulated authority, discards pages that were quietly converting, and introduces migration risk — all of which are worth accepting when the architecture is genuinely the constraint, and none of which are worth accepting otherwise.
Four signals that justify a rebuild
- The architecture blocks what the business now needs. You cannot add the section, the language or the product type without fighting the system.
- The platform is unsupported or the dependency chain is too far behind to update incrementally. At that point a rebuild is cheaper than the upgrade path.
- The technical foundation makes search and AI visibility impossible — content that only renders after hydration, a URL structure that cannot be corrected, no way to control canonicals.
- Maintenance cost has exceeded what a rebuild would cost. This is arithmetic, not judgment: add up twelve months of patching.
Six signals that do not
- It looks dated. Aesthetic fatigue is real and it is not a commercial argument.
- A new marketing lead arrived. The most common unstated reason for a redesign, and the least defensible.
- A competitor launched something. Their redesign may also have been a mistake.
- Conversion is poor. Usually a message or funnel problem, and fixable on the current site for far less.
- It is slow. Almost always images, third-party tags and render strategy — none of which require a rebuild.
- Rankings dropped. Diagnose the cause first. Rebuilding an unexamined ranking problem frequently makes it worse.
The test worth running first
Write down what the site fails at, in commercial terms, and what evidence you have. If the list is one or two items and the evidence is analytics rather than opinion, you have a fix. If the list is long and every item traces back to the same structural cause, you have a rebuild.
If you cannot write the list, that is the most useful finding available. A redesign brief without it produces a site that looks different and performs the same, which is the outcome most redesigns actually deliver.
If you do rebuild
Establish the baseline first — what ranks, what converts, what earns links — and keep it regardless of who does the work. Then map every URL before build starts. Those two steps are what separate a redesign that gains ground from one that spends money to lose it.