Web Development
Website Redesign: When to Rebuild, When to Improve What You Have
A redesign is the right move when the current site cannot say the offer clearly, cannot be edited without breaking, or cannot meet basic speed and mobile standards no matter how you patch it. It is the wrong move when leads are dying on a confusing form, a slow image, or a headline that still talks about you instead of the buyer. Rebuilds feel productive. Diagnosis is cheaper.
Separate “it looks old” from “it fails at the job”
Owners often notice the look first because look is visible in a screenshot. Users notice something else: can I find the phone number, does the menu work with a thumb, does the page finish loading before I give up. A 2018 layout with clear service pages and a working form can still outperform a 2026 aesthetic that hides the next step behind a carousel. Treat appearance as one input, not the verdict.
Before you hire anyone, export a month of behaviour if you have analytics. You do not need a fancy dashboard. You need to know which pages people land on, which pages they leave, and whether contact or checkout is even in the path. If you have no analytics, install it and wait for a real sample rather than guessing from your own browsing on a fast office connection.
A diagnosis checklist you can run this week
Work through these questions with the person who answers enquiries, not only with the person who pays the invoices. Those two people often disagree about what is wrong, and the front-line view is usually more accurate.
- Can a stranger name your offer after ten seconds on the home page, on a phone?
- Is the primary action visible without scrolling on a typical mobile screen?
- Do form submissions arrive, get answered, and get tagged by source?
- Are key pages slow because of huge images, chat widgets, or tag managers?
- Does the CMS fight you when you change a price or a service area?
- Are URLs, titles, and headings so messy that search engines and humans both guess?
- Is the visual language so inconsistent that the company looks accidental?
If the first three fail, you likely have a conversion and operations problem. That is the work in website conversion optimization, and it does not require new branding. If four fails, start with website performance optimization. If five fails, you may need a better editing model, which might be a redesign of the CMS, not of the public look. Only six and seven, especially together with a changed offer, make a full visual rebuild the honest recommendation.
Three paths, not one project named “redesign”
Path A: Keep the site, fix the path
Use this when the information architecture is basically right and the brand is still accurate. Rewrite the home-page lead. Shorten the form. Move testimonials next to the action instead of on a dedicated page nobody opens. Compress images. Remove scripts you cannot name. This path is unglamorous and often produces the first measurable improvement.
Path B: Reskin on the same structure
Use this when the pages and URLs should stay, but the visual system is dated or inconsistent. You keep the sitemap, the CMS content types, and the forms. You change type, color, spacing, and component styles. The risk is treating a reskin as a chance to “just tweak the nav,” which quietly becomes a rebuild. Write a freeze list: URLs that must not change, forms that must not be redesigned until after launch, and third-party embeds that stay as they are.
Path C: Rebuild
Use this when the offer has changed, the CMS is a trap, or the templates cannot represent the content you now sell. A firm that used to sell one service and now sells three packages with different intake questions is a rebuild candidate. So is a site that was assembled from five themes and two page builders. Rebuilds should still reuse what is true: copy that works, photos that are real, and any URL that already ranks or is printed on paper.
When a rebuild is justified even if the site “works”
Working is not the same as fit for purpose. Accessibility failures, mixed HTTP assets, plugin debt that nobody can update, or a theme abandoned by its author are operational reasons to start over. So is a merger, a rename, or a move from a personal brand to a company brand. In those cases, delaying a rebuild means paying twice: once to keep the old thing alive, and again when it finally collapses during an update.
- Security updates cannot be applied without breaking the layout.
- The only person who understood the theme has left.
- Mobile navigation was never designed; it was squeezed.
- Legal pages, cookie behaviour, or form storage no longer match how you operate.
- You need structured content (locations, staff, SKUs) and the current pages are giant blobs of HTML.
How to brief a redesign without inviting scope creep
Write a one-page brief with four sections: what must stay (URLs, integrations, tone), what must change (offer, audience, proof), what “done” means on a phone, and what is explicitly out of scope. Attach three competitor or peer sites you respect and one sentence about why — “their service pages state price ranges” is useful; “it looks premium” is not. Include who will supply photos and who will approve copy. Design-by-committee is how redesigns slip three months.
| In the brief | Leave out of the first phase |
|---|---|
| Primary conversion and who handles it | A blog strategy you will not staff |
| List of URLs to keep or redirect | A custom illustration system |
| Devices and browsers you actually see in analytics | Support for obsolete browsers with no traffic |
| Integrations that must survive (forms, booking, chat) | A new CRM migration in the same sprint |
Measure the new site against the old job
If the old site’s job was booked calls, the new site should be judged on booked calls, not on whether the team likes the hero video. Keep the same analytics events where you can. Launch with redirects mapped, not “we will fix SEO later.” A redesign that drops the phone number from the header because it cluttered the new grid has already failed a common business test.
Scriplit can help you choose among the three paths if you send the current URL, the offer sentence, and the diagnosis checklist answers via the web development contact form. A rebuild is sometimes the right recommendation. Often it is not, and you should hear that before you archive a site that only needed a faster contact path.