Here are the 8 signs you need a new website, at a glance:
- It's slow,
- It's awkward on a phone
- Traffic is drifting down
- Visitors arrive but nobody enquires
- You can't update it yourself
- It shows a security warning or something's visibly broken
- It no longer matches your business
- You're invisible when people ask Google or an AI assistant about what you do.
Not every sign means a full rebuild. Some of them are usually fixable in a week.
Websites rarely fail loudly. There's no alarm when a page loads too slowly or a contact form quietly stops sending. It leaks instead: slightly lower rankings this month, a few fewer enquiries than last quarter, a visitor who leaves without quite knowing why. Each sign below comes with a test you can run yourself, free, in about two minutes.
The 8 signs
1. It's slow, and Google agrees
What it looks like: Pages take a beat too long to load, especially on mobile. You might not notice on your own fast office Wi-Fi.
Why it costs you: Google's 2016 "Need for Mobile Speed" research found that 53% of mobile visitors abandon a page that takes longer than three seconds to load. Slower sites also rank behind faster competitors for the same searches, since page experience is a ranking factor.
The 2-minute test: Run your homepage through Google PageSpeed Insights. Look at the Largest Contentful Paint (LCP) score. Anything over 2.5 seconds on mobile is outside Google's "good" threshold.
2. It's awkward on a phone
What it looks like: Text too small to read without zooming, buttons too close together to tap accurately, a menu that covers half the screen.
Why it costs you: Google indexes and ranks primarily using the mobile version of your site, not the desktop one. If mobile is the afterthought, so is your ranking, and so is the experience for the majority of your visitors, who are almost certainly arriving on a phone.
The 2-minute test: Open your own site on your own phone. Try to find your phone number and submit your contact form without zooming in once.
3. Traffic is drifting down
What it looks like: Nothing dramatic. Just a little less each month, easy to miss without checking.
Why it costs you: A slow decline is often the compounding effect of the other signs on this list: falling rankings from poor speed, content that's stopped matching what people search for, or competitors publishing more than you are.
The 2-minute test: Open Google Search Console and look at your 12-month clicks trend. A steady downward line is worth investigating even if no single month looks alarming.
4. Visitors come but nobody enquires
What it looks like: Analytics show reasonable traffic. The phone doesn't ring and the contact form stays quiet.
Why it costs you: Traffic without conversion usually means the site isn't answering the question a visitor actually came with, whether that's price, trust, or simply how to get in touch. Fixing this is often cheaper than getting more traffic to a page that wasn't converting anyway.
The 2-minute test: Compare monthly sessions in your analytics tool against form submissions or call volume for the same period. If the ratio looks low relative to your industry, that gap is the problem to solve before spending more on getting visitors there in the first place.
5. You can't update it yourself
What it looks like: Every small change, a new price, an updated opening hour, a fixed typo, means an email to a developer and a wait.
Why it costs you: Beyond the delay, this usually comes with per-change fees that add up over a year, and it means your site quietly falls out of date because updating it is a hassle rather than a five-minute task.
The 2-minute test: Try to change a single headline yourself, right now. If you can't find where, or you're not confident enough to try without help, that's the CMS bottleneck showing itself directly.
6. It shows "not secure" or looks visibly broken
What it looks like: A browser warning next to your address bar, an image that never loads, a link that leads nowhere.
Why it costs you: A missing SSL certificate or a visibly broken page does more reputational damage in three seconds than almost anything else on this list. It tells a new visitor, correctly, that nobody's been paying attention.
The 2-minute test: Look for the padlock icon next to your URL in the address bar. Click through your main navigation and your most important two or three pages, checking for broken images or dead links as you go.
7. It no longer matches your business
What it looks like: Services you no longer offer, prices that changed two years ago, a team page missing half the current team, branding that doesn't match what's on your business card.
Why it costs you: A widely cited 2006 study from Carleton University found that visitors form an opinion of a website within 50 milliseconds, long before they've read a word. A site that visibly doesn't match the business behind it undermines trust before the content gets a chance to make the case.
The 2-minute test: Read your homepage and services pages as if you were a new customer who knows nothing about your business. Flag anything that's factually out of date.
8. You're invisible in AI and search answers
What it looks like: Search your own service and city, or ask ChatGPT or Google's AI overview the same question, and your business doesn't come up at all.
Why it costs you: More people now research purchases through AI answers as well as Google, and both rely on clear, well-structured content to understand what you offer and who it's for.
The 2-minute test: Ask an AI assistant a question a prospective customer might genuinely ask, phrased naturally, like "[your service] in [your city]." Then run the same search on Google and check whether you appear in the AI overview.
Which signs mean a rebuild, and which are quick fixes
Sign | Usually fixable in a week | Structural (needs real work) | Often rebuild territory |
|---|---|---|---|
Slow pages | Sometimes (image compression, caching) | If the platform itself is the bottleneck | If several signs fail together |
Awkward on mobile | Rarely | Usually | Often |
Traffic drifting down | Sometimes (content refresh) | Usually needs deeper SEO work | Not usually |
Visitors don't enquire | Sometimes (CTA and copy fixes) | If it's a structural conversion path issue | Not usually |
Can't update it yourself | Not usually | Not usually | Usually, since it's a platform/CMS problem |
"Not secure" or broken elements | Yes, almost always | Not usually | Not usually |
Content out of date | Yes | Not usually | Not usually |
Invisible in AI/search | Sometimes (schema, structure) | Often | If content architecture is the root cause |
A redesign isn't automatically the answer. If you're failing one or two signs in the fixable column, fix those first. A rebuild that doesn't address the structural causes on this list is just a new coat of paint on the same problems, and they'll come back within a year.
How often should a website be redesigned?
There's no fixed cycle, whatever a "redesign every three years" rule of thumb implies. Three to five years is a common average in practice, but it's a symptom of how long sites tend to accumulate the signs above, not a deadline. A six-year-old site that passes every test in this post doesn't need touching. A one-year-old site that fails four of them needs attention now. Redesign when the tests fail, not when the calendar says to.
What a redesign should fix, and what it should keep
A proper redesign fixes the failed tests: rebuilds the CMS bottleneck, restructures for mobile and speed, and closes the conversion gap. It should also deliberately keep what's already working: your existing rankings, your URL structure where it still makes sense, and any content that's genuinely performing. Losing those in the process is the most common way a redesign backfires. If your page addresses have to change, redirect each old one to its new one. That keeps the rankings you've earned. Our website migration cost guide explains what that costs.
Failed more than two of these tests?
That's usually the point where fixing things one at a time stops being the efficient path.
Book a discovery call. We'll tell you which fixes are worth doing on the current site, and whether a rebuild actually pays for itself.
