Every week someone asks us the same question in different words: “Is our website actually bad, or are we just bored of looking at it?”
It is a fair question. A redesign is one of the more expensive line items a small or mid-sized business will approve in a given year, and “it feels dated” is not a business case. So before we quote anything, we run the same diagnostic pass on every site that lands in our inbox. Some of it is tooling. A lot of it is just knowing where the bodies are buried in a five-year-old WordPress install.
This post is that diagnostic, written out. Eleven signs you need a website redesign, the thresholds we actually use, and at the end a scoring sheet you can fill in yourself in about five minutes to decide between a light refresh and a full rebuild.
What We Audit First (Before We Even Look at the Design)
Here is the uncomfortable part: the visual design is usually the last thing we check. A beautiful site that loads in six seconds on 4G and leaks form submissions is worth less than an ugly one that converts. Our first pass, roughly thirty minutes, covers:
- Core Web Vitals from field data (real users), not just lab scores
- Mobile versus desktop behaviour in Google Analytics 4: bounce rate, engagement time, conversion rate split
- Every form on the site, submitted for real, with the inbox checked
- The plugin list, the theme, the PHP version and the last update dates
- Google Search Console: indexing errors, impressions trend over 16 months, page experience
- Content freshness and whether the site reflects what the business actually sells today
Then, and only then, the design. Below is what each of those checks turns up, and what the number has to be before we call it a red flag.
The 11 Signs You Need a Website Redesign
1. Your Mobile Bounce Rate Is Far Worse Than Desktop
Open GA4, segment by device, and compare engagement rate on mobile versus desktop for the last 90 days. A small gap is normal. Mobile traffic is more casual, more top-of-funnel, more accidental.
A gap of 15 percentage points or more is not casual browsing. That is a usability failure. We see it most often on sites built between 2016 and 2020 that were technically “responsive” but designed desktop-first, then squeezed down. The tell-tale symptoms:
- Body text under 16px, forcing users to pinch and zoom
- Tap targets smaller than 44 by 44 pixels, sitting too close together
- Tables or pricing grids that overflow horizontally
- Sticky headers eating 25% of the viewport on a phone
- Hero images that push the actual message below the fold
With mobile now driving the majority of first touches for most local and B2C businesses, and a healthy share of B2B research, this one on its own can justify a rebuild.
2. You Are Failing Core Web Vitals in Field Data
Lab tools lie a little. They run on a simulated device in a data centre. What matters is the Chrome User Experience Report data, which you can see in Search Console under Page Experience, or in PageSpeed Insights at the top of the report if your site has enough traffic.
| Metric | Good | Needs work | Red flag |
|---|---|---|---|
| LCP (largest contentful paint) | Under 2.5s | 2.5s to 4.0s | Over 4.0s |
| INP (interaction to next paint) | Under 200ms | 200ms to 500ms | Over 500ms |
| CLS (cumulative layout shift) | Under 0.1 | 0.1 to 0.25 | Over 0.25 |
The important nuance: failing Core Web Vitals does not always mean you need a redesign. Sometimes it means you need better hosting, image compression and a caching layer. We have taken sites from 6s to 1.9s LCP without touching a single pixel of the design.
It becomes a redesign trigger when the slowness is structural: a page builder generating 400KB of inline CSS, a slider plugin loading jQuery UI on every page, or a theme that ships twelve Google Font weights. At that point you are paying rent on someone else’s bad architecture.
3. Your Forms Are Broken, and Nobody Noticed
This is the single most common finding in our audits, and the one that makes clients go quiet on the call.
Forms break silently. An SMTP plugin gets deprecated, a hosting provider blocks the PHP mail function, a reCAPTCHA key expires, a Gmail filter sends everything to spam, or the person whose address receives the notifications left the company eighteen months ago. The form still shows “Thanks, we will be in touch” and nothing arrives anywhere.
Go and test every single form on your site right now. Contact, quote request, newsletter, download gate, booking. Submit them with a real address and confirm the message lands. Also check:
- Are submissions stored in the database as a backup, or only emailed?
- Does the form work on iOS Safari, not just Chrome on your laptop?
- How many fields does it have? Anything over six for a first contact is costing you conversions.
- Is there a visible error state when validation fails, or does the page just refuse to submit?
4. Plugin Bloat Has Taken Over
A typical healthy WordPress business site runs somewhere between 12 and 20 plugins. We regularly open installs with 45 or more. The problem is not the number itself, it is what the number represents: years of patching gaps instead of fixing the underlying build.
What we look for in the plugin list:
| Finding | What it usually means |
|---|---|
| Plugins not updated in 12+ months | Abandoned by the developer, a security liability |
| Three plugins doing overlapping jobs | Nobody owns the stack, things were added ad hoc |
| A page builder plus a second builder plus Gutenberg | Multiple redesign attempts layered on top of each other |
| Nulled or unlicensed premium plugins | No updates, frequently malware. Rebuild, do not patch. |
| Theme with 20+ bundled plugins it depends on | You are locked in, migration cost grows every year |
Also check your PHP version in Tools, Site Health. If you are still on PHP 7.x in late 2026, you are running unsupported software on a public server.
5. Nobody on Your Team Can Update the Site
Ask your marketing person to add a new service page. Time it. If the honest answer is “I have to email the agency” or “it takes me two hours and it never looks right,” your site has an editorial problem, and editorial problems compound.
Sites that are hard to update do not get updated. Content goes stale, the blog stops in 2023, the team page shows people who left, and your organic traffic quietly slides. We treat editor friction as a top-three redesign trigger, because it predicts every other decay signal on this list.
6. Your Design Is Carrying Visible Date Stamps
Some patterns put a year stamp on your business the moment a visitor lands. Not because they are ugly in the abstract, but because the entire web moved on and your visitors have been trained by thousands of other sites.
- A full-width rotating hero slider with three messages nobody reads past slide one
- Narrow boxed layouts with heavy drop shadows and a visible content “frame”
- Body copy at 13px or 14px on desktop
- Obvious stock photography: the handshake, the headset call-centre woman, the diverse team laughing at a laptop
- Gradient bevel buttons, hard shadows, skeuomorphic icons
- Splash or “enter site” pages, autoplaying audio or video with sound
- Social icons linking to a dormant X account and a Facebook page last posted on in 2021
- Content locked at 1140px wide on a 27 inch monitor with oceans of empty margin
None of these is fatal alone. Three or more together and visitors are making a judgement about your competence before they read a word.
7. Organic Traffic Has Been Sliding for Two or More Quarters
Pull 16 months of Search Console data and look at impressions, not just clicks. Clicks can drop for reasons outside your control, including AI overviews taking the top of the page. Impressions falling steadily means Google is showing you for fewer queries, which is a content and technical problem. See envydesign.co for their take.
Before blaming the design, check:
- Indexing: how many pages are indexed versus submitted? A large gap means crawl or quality issues.
- Thin pages: service pages with 150 words of copy do not rank in 2026, and they do not get cited by AI answer engines either.
- Site structure: if your key pages are four clicks deep with no internal links pointing at them, that is an information architecture problem a redesign fixes.
- Cannibalisation: six near-identical pages competing for the same term.
8. Your Site Does Not Describe What You Actually Sell
Businesses evolve faster than their websites. We regularly audit sites where the highest-margin service is buried in a dropdown, and a product the company stopped offering in 2022 still owns the homepage.
Quick test: open your homepage, read only the headline and subheadline, and ask whether a stranger could name your main offer and who it is for. If the answer is a vague phrase like “innovative solutions for modern business,” you have a positioning problem that no amount of new CSS will solve.
9. Conversion Rate Is Below Benchmark Despite Decent Traffic
Traffic without conversion is the most expensive kind of website. If you are getting a few thousand sessions a month and converting under 1% to any meaningful action, the funnel is leaking. A good illustration of it comes from a studio that gets this right.
What we look at:
- Is there a clear primary call to action above the fold, repeated down the page?
- How many clicks from landing to enquiry? Three or fewer is the target.
- Is pricing or at least a price range visible anywhere? Hiding it kills B2C and local conversion.
- Is there social proof near the decision point, or are all testimonials dumped on a separate page nobody visits?
- Scroll depth on key pages: if 70% of users never reach your CTA, the page is too long or too slow to get to the point.
10. Accessibility Is Not Even Partially Handled
This has moved from “nice to have” to a genuine risk area, especially for businesses selling into Europe where accessibility requirements now apply to a wide range of commercial digital services. Beyond compliance, poor accessibility directly costs you customers.
Run any free accessibility checker on your homepage and three inner pages. Red flags:
- Text contrast ratios below 4.5:1 (light grey on white is the classic offender)
- Images with no alt text, or alt text that reads “image1234.jpg”
- Forms with placeholder text instead of real labels
- Navigation that cannot be operated with a keyboard
- Missing heading hierarchy, with H1s used as decoration
11. The Site Is a Security Liability
The last check, and the one that most often converts a “maybe next year” into “start Monday.” Signs that your site is already compromised or close to it:
- No SSL, mixed content warnings, or an expired certificate
- Admin accounts belonging to agencies or freelancers you parted ways with
- WordPress core more than two minor versions behind
- No off-server backups, or backups nobody has ever tested restoring
- Spam pages indexed under your domain (search site:yourdomain.com and scroll to the end)
- Google Search Console security issues notifications you dismissed
The 5-Minute Self-Scoring Framework
Score each sign from 0 to 3. Be honest. Total out of 33.
- 0 = not an issue at all
- 1 = minor, slightly annoying
- 2 = clearly a problem, costing us something
- 3 = severe, actively losing us money or trust
| # | Sign | Your score (0-3) |
|---|---|---|
| 1 | Mobile experience and bounce gap | |
| 2 | Core Web Vitals / load speed | |
| 3 | Forms and lead capture | |
| 4 | Plugin bloat and technical debt | |
| 5 | Editor friction (can your team update it?) | |
| 6 | Dated visual design patterns | |
| 7 | Organic visibility trend | |
| 8 | Offer and positioning mismatch | |
| 9 | Conversion rate and funnel clarity | |
| 10 | Accessibility | |
| 11 | Security and maintenance state |
Reading Your Score
| Total | Verdict | What we would recommend |
|---|---|---|
| 0 to 8 | Maintain | No redesign. Put the budget into content, speed tuning and conversion testing. |
| 9 to 16 | Refresh | Keep the platform and structure. Restyle templates, rewrite key pages, fix forms, prune plugins. Typically 3 to 5 weeks. |
| 17 to 24 | Partial rebuild | New theme or block-based build, new information architecture, content migrated and reworked. Keep the domain history and URLs where possible. 6 to 10 weeks. |
| 25 to 33 | Full rebuild | Fresh build from the ground up, new design system, full content strategy, redirect map. 10 to 16 weeks depending on page count. |
One override rule: if you scored a 3 on security (sign 11) or on forms (sign 3), fix that this week regardless of your total. Those are bleeding wounds, not design decisions.
Refresh vs Rebuild: How We Actually Decide
The score gives you a direction. The final call usually comes down to three questions. Full details on https://summationx.com.
Is the foundation sound?
If the site runs on a well-maintained theme or a clean block-based build, with sensible templates and no nulled code, a refresh gets you 80% of the result for 35% of the cost. If it runs on an abandoned ThemeForest theme from 2017 with a page builder that has not shipped a major update in three years, every hour spent refreshing is an hour spent on something you will throw away.
Is the problem visual or structural?
Colours, typography, spacing, imagery and button styles are visual. Those are cheap to change. Navigation logic, URL structure, how content types relate to each other, how a new service page gets created: those are structural. Structural problems need a rebuild.
What changes in the business in the next 24 months?
If you are adding a second product line, expanding into another language, launching e-commerce or moving upmarket, build for that now. We have seen too many clients spend on a refresh in Q1 and come back needing a rebuild by Q4 because nobody asked what was coming.
What to Do Before You Brief an Agency
- Record your baseline. Screenshot your GA4 numbers, Search Console impressions and Core Web Vitals today. Without a baseline you cannot prove the redesign worked.
- Define one primary goal. More qualified enquiries, lower cost per lead, shorter sales cycle, fewer support calls. One, not five.
- Export your top 50 pages by traffic and by conversion. These must survive the migration with their URLs intact or redirected properly. This is where most redesigns lose organic traffic.
- List who owns content after launch. If nobody owns it, you will be back here in three years.
- Agree the measurement window. Judge the result at 90 days post-launch, not week two when rankings are still settling.
The Honest Summary
Most of the signs you need a website redesign are measurable. Mobile engagement gaps, LCP over four seconds, dead forms, forty plugins, flat impressions, a 0.4% conversion rate. If you can put numbers next to three or four of the eleven signs above, you do not have a taste problem, you have a performance problem, and the redesign pays for itself.
And if you score under 9? Keep the site. Spend the money on content and speed instead. We will happily tell you that, because the fastest way to lose a client is to sell them a rebuild they did not need.
Want a second opinion on your score? We run the full audit on your actual site, with the real numbers, and tell you honestly whether a refresh will do the job. Get in touch.
Frequently Asked Questions
How often should a business website be redesigned?
The old “every three years” rule is outdated. A well-built site on a maintained platform, with ongoing content and performance work, can run five to seven years before a full rebuild. What should happen continuously is iteration: new pages, updated copy, conversion improvements, speed tuning. Sites that are never touched are the ones that need dramatic, expensive redesigns.
Will a redesign hurt my SEO rankings?
It can, if it is handled badly. The main risks are changing URLs without redirects, losing on-page content during migration, dropping internal links, and accidentally blocking crawlers with a leftover staging robots.txt. Done properly, with a full redirect map and content preserved or improved, a redesign usually lifts rankings within 60 to 90 days because speed, structure and content depth all improve at once.
How much does a website redesign cost?
It depends almost entirely on page count, content work and functionality. A refresh of an existing WordPress site sits at the low end. A full rebuild with new design system, custom blocks, content strategy and migration sits considerably higher. The useful question is not the price but the payback: if your site generates leads, calculate what a 20% conversion improvement is worth annually and compare.
Should I redesign or just improve my site speed?
Start with speed if that is your only failing metric. Better hosting, image optimisation, caching and removing unused scripts often fix Core Web Vitals without any design change, at a fraction of the cost. Redesign when the slowness is baked into the theme or builder, or when speed is one of several problems on the list above.
Can I keep my existing content in a redesign?
Yes, and you usually should keep the pages that already rank and convert. We audit content before anything else: keep and improve, consolidate, or retire. Wholesale deletion of ranking pages is the single most common cause of post-launch traffic drops.
What is the difference between a website refresh and a rebuild?
A refresh keeps the existing platform, templates and URL structure, and changes the surface: styling, imagery, copy, calls to action, plus some technical cleanup. A rebuild replaces the underlying build, which lets you fix information architecture, performance and content modelling, but costs more and takes longer. The scoring table above is designed to tell you which one you need.