If you’re asking this question, chances are your site either “looks old,” or something has broken that you can’t fix yourself. In my consulting work, these are the two most common triggers I see. But the honest answer is that timing depends less on what year it is and more on what your data and your business are telling you.
The Short Answer: Every 2-3 Years
As a general rule of thumb, most businesses should expect to do a substantial refresh every two to three years, with a full platform rebuild every three to five years depending on your industry and growth trajectory. But in my experience working directly with clients, the calendar is rarely the real trigger… It’s usually a “vibe” shift, a business pivot, or a technical wall they’ve hit.
Website Refresh vs. Website Rebuild
I draw a clear line between these two for clients, because conflating them leads to wasted budget. A refresh is cosmetic, which means new colours, updated copy, refreshed imagery on your existing platform. But a rebuild means changing the underlying architecture, often the CMS itself.
That difference matters most when I see clients migrating off WordPress. A huge portion of my work involves moving clients from WordPress to Squarespace. (Specifically so they can edit their own site without calling a developer every time they need to change a headline.) When we’re already in there rebuilding the scaffold, it almost always makes sense to modernise the look as well. So what might start a technical migration becomes a full visual refresh too.
The Signals I Look For (Beyond “It Looks Old”)
“It looks old” is a valid reason to redesign, but it’s a subjective one. What I actually push clients toward is looking at their data before committing budget to a redesign, because gut feel and analytics don’t always 100% agree:
- Mobile conversion drop-off: if desktop visitors buy but mobile visitors bounce, that’s a UX problem no amount of “the design feels dated” intuition will catch on its own.
- A noticeable lack of engagement on your most important pages, even when traffic volume is stable or growing.
- A CMS that’s become a bottleneck — if the client is emailing me every time they need a text change, that’s a platform problem, not just a design one.
- Business repositioning: new services, new target market, or a rebrand that the current site no longer reflects.
- Site speed and technical debt, especially on older WordPress builds carrying years of plugin bloat.
When “The Business Has Changed” Is the Real Trigger
Sometimes none of the technical signals are flashing red, but the business itself has outgrown the site. This can look like new offerings, a shift in positioning, or simply a founder who’s evolved past the brand they launched with. This is a legitimate and common reason to redesign, and it’s one that data alone won’t surface. It requires the kind of conversation I have directly with clients about where the business is heading, not just where the analytics currently sit.
My Approach: Data-Informed, Not Vibe-Only
I tell clients that “it looks old” is a fine starting point for a conversation, but it shouldn’t be the only justification for spending the budget. Before recommending a redesign, I look at analytics — particularly what data about mobile users says — to confirm whether the instinct matches reality. If mobile users are dropping off before purchase, that’s concrete evidence a redesign (or at minimum a user experience overhaul) will pay for itself, rather than just refreshing the site because it feels due.
Next Steps: Getting a Proper Audit
Whether your trigger is an outdated look, a WordPress site that’s become unmanageable, or conversion numbers that don’t add up, the right first step is a proper audit before committing to a full rebuild. I offer this as a starting point for clients unsure whether they need a light refresh, a platform migration, or a ground-up redesign. I put consultation and audit firsts so the decision is based on evidence, not just a gut feeling that “it’s time.”
Get in touch with me for a consultation and build with intention, not guesswork.