In short
There is no fixed redesign interval. Review the website continuously and consider a substantial redesign when the offer, audience, brand, content model, user journeys, or technical foundation has materially changed. If the system still supports the business, targeted improvements are usually safer and cheaper than replacement.
There is no responsible two- or three-year rule
A website can be five years old and effective, or six months old and structurally wrong. Age is a weak proxy for fit.
The better operating model is continuous maintenance, measured improvement, and occasional structural redesign. Replace the system when its assumptions no longer match the business or users.
Signals that justify a substantial redesign
The offer or audience changed
New products, markets, business models, or buyer groups may no longer fit the existing architecture. If visitors must understand the old organisation before finding the new offer, editing individual pages will not solve the navigation problem.
The brand changed materially
A refreshed logo can be applied. New positioning, voice, evidence, or market category may require a different information hierarchy and page system.
Important journeys fail
Research, analytics, enquiries, and support show that users cannot find, compare, trust, or complete priority tasks. Redesign is justified when the failure spans several templates or the navigation model—not one button.
The CMS obstructs publishing
Editors rebuild layouts manually, duplicate shared information, depend on developers for ordinary work, or cannot manage languages and relationships safely. That is a system problem.
The technical foundation blocks quality
Unsupported software, fragile plugins, inaccessible components, poor performance, or an architecture nobody can maintain may make incremental repair more expensive than replacement.
The site cannot support required integrations
The business needs product data, accounts, CRM routing, booking, payments, or regional workflows the current system cannot represent reliably.
Problems that usually do not require a full redesign
- one underperforming landing page;
- outdated photography;
- weak copy on a few service pages;
- isolated accessibility defects;
- a slow hero image;
- a new campaign;
- an executive's visual boredom;
- a competitor launching something fashionable.
Fix the smallest layer that owns the problem. A full redesign concentrates cost and risk; use it when the problem is genuinely systemic.
Run a quarterly health review
Review four areas:
| Area | Questions |
|---|---|
| Business fit | Does the site represent the current offer, audience, and evidence? |
| User fit | Can priority users complete important tasks? |
| Editorial fit | Can the team publish accurately and consistently? |
| Technical fit | Is the system secure, accessible, fast, observable, and maintainable? |
Record issues, consequence, evidence, owner, and the smallest plausible intervention. Patterns across areas may form the case for redesign.
Improve between redesigns
A healthy website changes through:
- content corrections and consolidation;
- navigation and internal-link improvements;
- form and conversion fixes;
- component accessibility work;
- performance budgets;
- CMS refinements;
- new evidence and case studies;
- measured experiments;
- dependency and infrastructure maintenance.
Continuous improvement extends the system's useful life and produces better evidence for the next redesign.
When to begin planning
Start before the current site becomes an emergency. A serious redesign may require discovery, content, migration, design, development, QA, stakeholder review, and post-launch monitoring.
If a brand launch, product release, or contract expiry creates the deadline, work backward from content and technical dependencies—not merely design production.
Common questions
Is three years the normal website lifespan?
No. Three years is a common sales shorthand, not a technical rule. Lifespan depends on business change, architecture, maintenance, content, and whether the system can evolve.
Can regular updates avoid a redesign?
Often. Continuous content, component, performance, and CMS improvements can preserve a sound foundation. They cannot fix an architecture built for a business that no longer exists.
Does an old-looking website always need redesigning?
No. First determine whether visual perception damages trust or task completion. A focused visual and content refresh may work without replacing URLs, CMS, and components.
Should redesign happen all at once?
Not always. Components, content types, or sections can be replaced progressively when the architecture permits it and users receive a coherent experience throughout the transition.
What should be measured before redesigning?
Record search performance, conversions, priority journeys, content inventory, accessibility, Core Web Vitals, support themes, editor workflows, and qualitative user evidence.
