A redesign changes the website system. Conversion optimisation improves specific parts of a system that is basically sound. Choosing the wrong route can create unnecessary cost or leave a structural problem untouched.
The decision should be based on evidence: business direction, user behaviour, content architecture, technology, search equity and operational requirements. Visual preference matters, but it is not enough on its own.
Optimise when the foundation is sound
Conversion optimisation is usually appropriate when the offer is understood, the website is technically stable, relevant visitors already arrive and the main problems can be isolated. Examples include an unclear landing-page message, weak proof, confusing calls to action or unnecessary form friction.
Optimisation works best when there is enough evidence to form a hypothesis and observe whether the change improves the intended action. It can include messaging, page structure, UX, form design, performance and lead-response changes.
Redesign when the system creates the problem
A redesign is more responsible when the information architecture cannot support the current business, mobile behaviour is consistently poor, the content model blocks publishing, performance problems are structural, the technology is difficult to maintain or the brand position has materially changed.
If every improvement requires another workaround, the system may be consuming more effort than a controlled rebuild. The redesign should still have a commercial purpose; “make it modern” is a direction, not a measurable objective.
Review five evidence areas
1. Business and offer
Has the audience, offer or sales model changed? If the website represents an older business, local optimisation may not solve the mismatch. If the offer is stable, targeted changes may be enough.
2. User journey
Can a serious visitor understand the value, find relevant proof and complete the next action on mobile and desktop? Identify whether friction is local to a page or repeated throughout the website.
3. Content architecture
Can the website support service pages, authority content, case studies and internal links without duplication? A redesign may be needed when content types and URLs cannot grow responsibly.
4. Technology and operations
Review performance, accessibility, security, editing workflow, integrations, analytics and maintenance. The best-looking rebuild will underperform if the team cannot operate it.
5. Search and conversion evidence
Inspect indexed pages, landing queries, impressions, clicks, lead sources and page-level actions. A valuable URL should not be removed casually. An unproven page should not be preserved merely because it exists.
Protect what already works
A redesign needs a URL inventory, redirect map, metadata plan, internal-link review and analytics baseline. Retain useful content and search equity. Redirect only old URLs with a genuine replacement; unknown and malicious paths should remain meaningful 404 responses.
Record important forms and lead workflows before launch. Test them after migration. A website is not complete if it looks correct but loses attribution or enquiry delivery.
A practical decision matrix
- Choose optimisation when the issue is specific, the system is maintainable and useful traffic already reaches the right page.
- Choose redesign when problems repeat across structure, mobile, content, technology or brand direction.
- Choose a phased approach when an urgent conversion problem can be repaired before a larger rebuild.
- Choose diagnosis first when stakeholders disagree about the problem or evidence is incomplete.
Define the smallest useful scope
An optimisation phase might improve one high-intent journey and its form. A redesign phase might establish the new architecture, migrate priority pages and connect the lead workflow before secondary content is expanded.
Both routes need a clear audience, purpose, owner and measurement plan. Neither should become a collection of aesthetic changes without a decision model.
Test the decision before expanding it
For optimisation, choose one high-intent page and write down the hypothesis, evidence and intended action before making changes. For redesign, prototype the navigation, page model and most important mobile journey before the full build. Early testing should expose structural mistakes while they are still inexpensive to correct.
After launch, compare the new experience with the recorded baseline. Confirm that priority URLs resolve correctly, metadata and canonicals are present, forms reach the CRM, analytics preserve attribution and the team can publish without new workarounds. A visual review alone cannot confirm a successful migration.
Review Landing Pages, Funnels & Commerce for a focused conversion route, or Conversion-Focused Websites when the system requires rebuilding. If the choice is unclear, use the project brief to request a human-reviewed recommendation.
