Nobody waits for slow websites anymore. That’s not a dramatic claim. It’s backed by years of conversion data, Google’s own research, and the lived experience of anyone who’s abandoned a product page because it took four seconds to load on their phone. Speed stopped being a nice-to-have feature around 2018. Now it directly affects your search rankings, your bounce rate, your revenue per session, and whether Google considers your site worth showing people at all.
Yet the gap between sites that take performance seriously and sites that don’t remains enormous. Most small and medium businesses are still running on themes and plugins chosen for how they look, not how they perform. And they’re paying for it every single month in lost traffic they’ll never know they lost.
What Core Web Vitals Actually Measure
Google’s Core Web Vitals framework breaks page experience down into three signals: Largest Contentful Paint (LCP), Interaction to Next Paint (INP, which replaced First Input Delay in 2024), and Cumulative Layout Shift (CLS). Each one measures something a real user actually experiences, not an abstract technical benchmark.
LCP measures how long it takes for the main content element on a page to fully render. Google’s threshold for a good score is under 2.5 seconds. Most unoptimised WordPress or Shopify sites on shared hosting sit between 3.5 and 6 seconds. That’s the “poor” range, and it sends a clear signal to the algorithm.
INP replaced FID because FID only measured the delay before the browser registered the first interaction. INP measures the actual time from any user interaction to the visual response. It’s a harder metric to pass because it catches sluggish JavaScript execution mid-session, not just on first load.
CLS catches visual instability: images without declared dimensions that cause content to jump, fonts that swap and shift text blocks, ads that push layout around after load. Users experience this as the page feeling broken or unreliable. Google treats it as a quality signal.
The Real Causes of Slow Sites
When someone asks why their site is slow, the honest answer is almost always “multiple things compounding.” Rarely is there a single culprit. Uncompressed images are usually part of it. A 2MB hero image that hasn’t been converted to WebP and served at the right dimensions for mobile adds seconds by itself. But that’s often sitting alongside render-blocking JavaScript in the head, a theme loading five CSS files that could be a single minified bundle, Google Fonts making external requests on every page load, and a caching plugin that’s either misconfigured or fighting with the server-level cache.
Server response time (TTFB, Time to First Byte) is the layer most site owners ignore entirely. If your server is on crowded shared hosting and takes 800ms just to respond before the browser can render a single pixel, you’ve already lost a significant chunk of your performance budget. A good TTFB is under 200ms. Many sites on cheap hosting are triple that.
Third-party scripts deserve a special mention. Google Tag Manager loading twelve tags, a live chat widget, a cookie consent popup, a heatmap tool, a retargeting pixel, and a customer review widget. Each one adds requests. Each one blocks or delays rendering to some degree. Auditing your third-party script load and removing anything that isn’t earning its keep is one of the highest-impact things you can do without touching the site’s core code.
How Images Are Still Killing Performance in 2025
You’d think image optimisation would be a solved problem by now. It isn’t. The number of production sites still serving PNG screenshots as hero images, uploading 4000px wide photos into blog posts, and letting WordPress serve the full-resolution original to mobile users is frankly staggering.
Proper image handling in 2025 means WebP or AVIF format, srcset and sizes attributes on every img tag so the browser fetches the right size for the viewport, lazy loading for everything below the fold, and explicit width and height attributes on every image to prevent layout shift. For LCP images specifically, you want preload hints in the head so the browser fetches them ahead of the normal waterfall. Missing that step is one of the most common reasons sites fail LCP even after other optimisations.
CDN delivery matters here too. If your images are sitting on a server in London and a user loads your site from Manchester, the distance is negligible. But if they’re loading from Edinburgh or abroad, serving assets from a CDN edge node near the user shaves real milliseconds off every asset request. Cloudflare’s free tier alone handles this for most small sites.
JavaScript: The Performance Enemy You Invited In
Modern websites carry extraordinary amounts of JavaScript. Some of it is necessary. A lot of it isn’t. Slider plugins that could be replaced with CSS-only carousels. jQuery still being loaded by legacy plugins when native browser APIs handle the same tasks. Entire animation libraries included for a single fade effect.
The performance cost of JavaScript isn’t just download size. The browser has to parse it, compile it, and execute it. On lower-end Android devices, that process can take multiple seconds even for moderately sized scripts. The move toward lighter, module-based JavaScript and deferring non-critical scripts until after the initial paint is one of the most meaningful performance improvements you can make on a busy site.
Unused JavaScript is a particular problem on WordPress sites with many plugins. Each plugin often loads its scripts site-wide even when the feature is only needed on one page type. Load Asset Manager Pro or Asset CleanUp Pro to conditionally load scripts only where they’re needed. The improvement on pages that don’t need those plugins can be significant.
Caching Strategy: Layers Matter
Caching is often treated as “install a plugin and tick a box.” A proper caching strategy is more layered than that. Server-level caching (Nginx FastCGI cache, LiteSpeed Cache at the server level, or Varnish) handles the majority of traffic before PHP is even involved. Application-level caching via W3 Total Cache, WP Rocket, or LiteSpeed Cache plugin handles the WordPress layer. Object caching with Redis or Memcached stores database query results so repeat queries don’t hit MySQL each time. Browser caching via cache-control headers tells returning visitors to use local copies of static assets rather than re-downloading them.
Running two caching systems that conflict with each other is worse than running none. It’s a common mistake. Speed of Light and LiteSpeed Cache active simultaneously, for instance, creates unpredictable behaviour. Pick one approach at each layer and make sure they’re not stepping on each other.
Mobile Performance Is Its Own Discipline
Desktop scores and mobile scores are not the same thing. Google uses mobile-first indexing, meaning the mobile version of your site is what determines your rankings. Desktop scores of 90+ can coexist with mobile scores of 40 on the same URL. This happens when sites serve the same JavaScript payload to all devices, when images aren’t properly sized for small viewports, or when touch targets are too small and trigger accessibility-related performance penalties.
Testing in Google PageSpeed Insights is essential but it uses a simulated mid-range Android device, not a flagship phone. That’s intentional. It approximates the experience of a real-world user on average hardware with an average mobile connection. Your site needs to perform well on that device, not just on the latest iPhone on a gigabit connection.
Performance Audits: What a Proper One Covers
A real performance audit isn’t running Lighthouse and reading out the suggestions. It covers TTFB and server configuration, resource waterfall analysis, render-blocking resources, Critical CSS identification, JavaScript bundle analysis, image delivery across breakpoints, caching headers, third-party script impact, font loading strategy, and Core Web Vitals field data from CrUX (real user data, not lab data). Lab data tells you what’s possible. Field data tells you what’s actually happening for your users.
This kind of audit requires experience and proper tooling: WebPageTest for waterfall analysis, GTmetrix, CrUX data in Search Console, and sometimes Chrome DevTools performance profiling for JavaScript execution traces. An agency offering this properly, like a specialist website speed optimisation UK service, is doing substantially more than running a free tool and generating a PDF report.
The Business Impact: What the Numbers Say
Walmart found that every one-second improvement in page load time increased conversions by 2%. Google found that 53% of mobile sessions are abandoned if a page takes longer than three seconds to load. Deloitte research showed that a 0.1 second improvement in site speed lifted retail conversion rates by 8.4%.
These aren’t edge cases from high-traffic enterprise sites. They apply at any scale. A local service business getting 500 visitors a month and converting at 3% would convert at roughly 3.25% after proper speed work. Over a year, that’s meaningful new revenue from the same traffic, without spending anything extra on ads or content.
Speed and SEO: The Relationship Isn’t Indirect
Some people still treat speed as separate from SEO. It isn’t. Page experience signals including Core Web Vitals are a confirmed ranking factor. Sites with poor CWV scores competing against sites with good CWV scores, all other things equal, lose. The advantage isn’t enormous but it’s real and measurable in competitive niches.
Beyond direct ranking signals, speed affects crawl budget. Googlebot doesn’t wait indefinitely for slow pages to respond. Faster sites get more pages crawled per session. For large e-commerce or content sites, that matters for how quickly new content gets indexed.
Speed also affects the behavioural signals Google infers. Lower bounce rate, longer session duration, more pages per session. These don’t directly tell Google your content is better, but they correlate with content quality at scale in ways that inform how Google ranks your site over time. Visiting Webranko gives you a clear picture of how this kind of technical SEO work connects to measurable business outcomes across different industries.
Where to Start If Your Site Is Slow
Run PageSpeed Insights and CrUX data from Search Console side by side. Look at your field data LCP, INP, and CLS. Check your TTFB. Audit your images with Screaming Frog or similar. List every third-party script loading on your site and cut anything that isn’t critical. Start with the changes that cost nothing: image compression, enabling gzip or Brotli compression on the server, setting proper cache headers, preloading your LCP image.
If you’ve done all of that and the scores are still poor, you’re looking at deeper work: server migration, theme refactoring, or a full technical audit. That’s when professional help pays for itself quickly.
Speed isn’t a luxury feature. It’s the baseline your site needs to compete. Everything else you do to grow your traffic, whether that’s content, ads, or backlinks, works better when the foundation performs.
