Page Speed OptimisationAboutContact
PILLAR
Page speed is the tax most sites don’t know they’re paying. Every extra second of load time costs around 7% of conversions on mobile, where 60%+ of your traffic now lives. Competitors who load in 2 seconds take the revenue that sites loading in 6 seconds bounce away.
Core Web Vitals, LCP, INP, CLS, aren’t Google rubric items. They’re revenue fences. A 0.5s LCP improvement on one fashion client recovered £180k a month. Speed is simultaneously the easiest CRO lever to measure and the hardest for most teams to actually ship, because it touches everything: image formats, font loading, JS bundles, third-party scripts, CDN choice, and server response.
Every post in this pillar tackles one practitioner fix: image pipelines, font display strategies, JS defer patterns, critical CSS, third-party script audits, and the monitoring you need to stop regressions landing unnoticed.
REVENUE IMPACT
Page speed is the time between a visitor clicking a link and being able to use the page they landed on. Two numbers matter for conversion: time to first paint (when does anything appear?) and time to interactive (when can the visitor actually click?). On mobile networks the second number is what costs you money.
The conversion math is uncomfortably linear. Each one-second improvement in mobile page load increases mobile conversion rate by ~7% (Google research, replicated in Akamai’s RetailMeNot study and Deloitte’s “Milliseconds Make Millions” 2020 report). Compounded across every paid-traffic visitor, that’s the difference between paid acquisition that breaks even and paid acquisition that compounds.
A real example: BeeFRIENDLY, a UK Shopify supplement store, reduced largest contentful paint by 2.24 seconds. Revenue went from $48K to $1.45M in the following 12 months. The speed improvement didn’t cause all of that growth — but it removed the constraint that was capping every acquisition channel.
The 2026 reality: Google now uses Core Web Vitals (LCP, INP, CLS) as a direct ranking signal. Slow pages are not only worse for the visitor — they’re harder to rank, more expensive to advertise on (Quality Score penalty), and harder to retain on mobile.
CORE WEB VITALS
Google grades pages on three Core Web Vitals. Hitting all three “Good” thresholds is the bar.
| Metric | What it measures | Good | Needs improvement | Poor |
|---|---|---|---|---|
| LCP (Largest Contentful Paint) | Time until the largest visible element renders | ≤ 2.5s | 2.5s–4.0s | > 4.0s |
| INP (Interaction to Next Paint) | Worst response time across all user interactions on a page | ≤ 200ms | 200ms–500ms | > 500ms |
| CLS (Cumulative Layout Shift) | Sum of layout shifts during the page session (lower is better) | ≤ 0.1 | 0.1–0.25 | > 0.25 |
INP replaced FID (First Input Delay) as a Core Web Vital in March 2024. FID only measured the first interaction; INP measures the worst interaction across the visitor’s whole session, which is a meaningfully harder bar.
The 75th-percentile rule: Google judges these metrics at the 75th percentile of your real-user traffic, not your lab tests. A page that scores 95 in Lighthouse but fails on slow Android devices will still be marked poor.
LCP FIXES
LCP is usually a hero-image problem or a server-response-time problem. The five fixes that move LCP the most:
<link rel="preload" as="image" href="hero.webp" fetchpriority="high"> in the document head. This tells the browser to start downloading the hero image before it parses the rest of the HTML.Real example: BeeFRIENDLY’s 2.24-second LCP reduction came from (1) WebP conversion (-700ms), (2) preload + fetchpriority (-540ms), (3) deferred review-app JS (-620ms), and (4) Cloudflare APO image-optimised cache (-380ms).
INP FIXES
INP is a JavaScript main-thread problem. When the visitor taps a button, the browser has to wait for the main thread to finish whatever it’s doing — running tracking pixels, hydrating React components, processing analytics — before it can respond.
The four moves that drop INP the fastest:
setTimeout(0), requestIdleCallback, or scheduler.yield where available.DOMContentLoaded.A site with a Klaviyo popup + a Yotpo review app + Meta Pixel + Hotjar can easily hit 600ms+ INP on a mid-range Android. Killing or deferring two of those four usually fixes the page.
CLS FIXES
CLS happens when something on the page moves after it first appears. The four fixes:
width: 100%, the height attribute lets the browser reserve the slot. Aspect-ratio CSS also works (aspect-ratio: 16/9).min-height before the ad loads.<link rel="preload" as="font" type="font/woff2" crossorigin> mostly eliminates this.position: fixed, not in the document flow.CLS is the cheapest of the three to fix and the one most sites still get wrong.
IMAGE OPTIMISATION
Images are typically 60–80% of a page’s transfer weight. Get this right and you’ve solved half the speed problem.
<picture> with <source type="image/avif"> and <source type="image/webp">.loading="lazy" to every image below the fold. Add loading="eager" and fetchpriority="high" to the LCP image only.alt (helps accessibility + SEO), explicit width/height (helps CLS).WordPress shortcut: install ShortPixel or EWWW Image Optimizer plus a CDN plugin (Jetpack Site Accelerator, Cloudflare). Three plugins, 70% of a hand-tuned image pipeline.
Shopify shortcut: Shopify’s CDN already auto-converts to WebP. The opportunity is image compression at upload time (use TinyIMG, Crush.pics, or Loading Speed Optimizer) and srcset for hero images that aren’t auto-handled.
TOOLS
| Tool | Best for | Free? |
|---|---|---|
| PageSpeed Insights | Single-page audit + Core Web Vitals lab + field data | Free |
| WebPageTest | Detailed waterfall, multi-location, multi-browser, multi-device testing | Free + paid |
| Lighthouse (Chrome DevTools) | Local audit during development | Free |
| Chrome DevTools Performance panel | Frame-by-frame main-thread analysis, INP debugging | Free |
| GTmetrix | Easier-to-read reports for non-engineers, video playback of load | Free + paid |
| Calibre | Continuous monitoring, regression alerts, lighthouse-CI integration | Paid (~$70/mo) |
| SpeedCurve | Enterprise monitoring, RUM + synthetic correlation | Paid (~$200/mo) |
| CrUX (Chrome User Experience Report) | Real-world field data for any public URL | Free (via API) |
| Search Console Core Web Vitals report | Site-wide CWV pass/fail breakdown, mobile vs desktop | Free |
Start free. PageSpeed Insights + Search Console + Chrome DevTools covers 95% of optimisation work. Pay for Calibre or SpeedCurve only when you have multiple developers shipping daily and need regression alerting.
For automated optimisation: Cloudflare APO ($5/mo for WordPress), Cloudflare Argo Smart Routing, Vercel Speed Insights, and platform-native tools like Shopify’s Online Store Speed report — all useful, none a substitute for an engineer who understands the trade-offs.
CDN + CACHING
What is a CDN? A Content Delivery Network is a network of edge servers that cache copies of your site’s static assets (images, CSS, JS, fonts) close to your visitors. A UK visitor hitting a US-hosted site adds 80–120ms of round-trip latency for every asset request. A CDN delivers those assets from a London edge server — 5–15ms. For an image-heavy ecommerce page, that’s a 30–50% LCP improvement on its own.
Cache-Control: public, max-age=31536000, immutable on hashed URLs.What to automate: image optimisation (CDN-side), CSS/JS minification (build step), cache purging on publish (webhook). What to do by hand: critical CSS extraction (per-template), third-party script audit (quarterly), and INP profiling (whenever a new app gets installed).
PLATFORM PLAYBOOKS
Each platform has different speed traps.
image_url filters out of the loop — use the asset CDN’s resize syntax (?width=) and srcset.after_dom_loaded or onload.SEO IMPACT
Yes directly and indirectly.
Directly: Core Web Vitals (LCP, INP, CLS) are part of Google’s page experience ranking signal. Pages that fail any of the three at the 75th percentile lose ranking in mobile search results. Google has confirmed this in their Search Central blog (May 2021 launch, March 2024 INP transition).
Indirectly, the bigger lift:
The 2.5s LCP / 200ms INP / 0.1 CLS triple is the bar. Hit all three at the 75th percentile of your real-user traffic and the SEO penalty is gone. Then the conversion math takes over.
FAQ
PageSpeed Insights for the single-page audit, Chrome DevTools Performance panel for INP debugging, and Search Console’s Core Web Vitals report for site-wide field data. WebPageTest if you need multi-location or multi-device testing. All four are free.
A CDN is a network of edge servers that cache copies of your site’s static assets close to your visitors. Instead of every request travelling to your origin server, the CDN serves assets from a geographically nearby edge — typically 5–15ms vs 80–200ms. The biggest CDNs (Cloudflare, Fastly, Bunny.net, CloudFront) cover the entire world from hundreds of edge locations.
Convert to AVIF (best) or WebP (universal); add responsive srcset for different viewport sizes; lazy-load every image below the fold with loading="lazy"; serve all images from an image CDN that handles dynamic resize and format conversion. WordPress: ShortPixel or EWWW. Shopify: TinyIMG or Crush.pics.
Yes. Core Web Vitals (LCP, INP, CLS) are a direct ranking signal in Google mobile search. Slow pages also lose crawl budget, raise bounce rate, and reduce AdSense Quality Score (which raises CPC).
WP Rocket for the easiest all-in-one (page cache + lazy load + critical CSS + JS defer, ~$59/year). LiteSpeed Cache for the fastest free option (requires a LiteSpeed-compatible host). Pair either with Cloudflare APO ($5/month) for HTML edge cache.
Three highest-leverage moves: (1) audit and defer third-party scripts (review apps, chat, popups, analytics); (2) move to a CDN with image optimisation; (3) optimise the hero image — preload, WebP/AVIF, explicit dimensions, fetchpriority="high". Most ecommerce stores get a 1–2 second LCP improvement from those three changes alone.
Cloudflare APO (automatic for WordPress and Shopify), Vercel Speed Insights, Calibre, SpeedCurve, NitroPack, and Shopify’s Online Store Speed report. For ongoing regression alerting, Calibre or SpeedCurve. For one-shot auto-optimisation, Cloudflare APO or NitroPack.
GoGoChimp (Glasgow) — 13 years of operator-led CRO and page-speed engagements, including the BeeFRIENDLY 2.24-second LCP reduction that took the store from $48K to $1.45M in 12 months. Other reputable UK firms include Conversion.com (London) and Convertize (London). For one-shot audits, GTmetrix Pro or PageSpeed.io.
INP (Interaction to Next Paint) measures the worst response time across all user interactions on a page. FID (First Input Delay) only measured the first interaction. Replaced in March 2024 because the first interaction was easy to game (defer everything until the first click). INP is a meaningfully harder bar to clear.
LCP ≤ 2.5 seconds, INP ≤ 200 milliseconds, CLS ≤ 0.1 — all measured at the 75th percentile of your real-user traffic (not your lab tests). Hit those three thresholds and Google considers your page experience “Good.” Conversion-wise, faster is always better, but the marginal return drops off below ~1.5s LCP on mobile.
FREE PAGE-SPEED AUDIT
Free 15-minute call. We’ll run your site through PageSpeed Insights + WebPageTest + Chrome DevTools, identify the three highest-leverage fixes, and quote you on the engineering work (or hand you the report so you can ship it yourself). No pitch — just the audit.
Get my free page-speed audit →RELATED
RELATED BLOG POSTS
COMING SOON
Image pipelines, font display strategies, JS defer patterns, critical CSS, third-party audits, and speed regression monitoring.
Run my free page speed audit →Three layers of intervention, in order. Image compression and WebP conversion (the single highest-leverage fix on most ecommerce sites, often cutting page weight by 80-90%). Render-blocking JavaScript audit and deferral (third-party tags are usually the worst offender). Server response time and CDN configuration. GoGoChimp's Speed Sprint is a £1,500 one-off engagement that hits all three across the highest-traffic pages on the site.
BeeFRIENDLY Skincare (Ezra Firestone brand) ran an annual revenue uplift from $48,000 to $1,447,225 after a 2.24-second page-speed reduction in 2017. Bounce rate fell from 82.04% to 38.4%. Per-visitor value went from $1.28 to $29.03. The intervention was theme-code edits, image compression, and WebP conversion. One operator, six months, ~30x revenue multiplier. Numbers held at least six months post-implementation.
For audit and measurement: PageSpeed Insights (Google's free Lighthouse-powered tool), GTmetrix, and WebPageTest. For caching: WP Rocket, W3 Total Cache, or LiteSpeed Cache. For image compression and WebP conversion: ShortPixel or Imagify. The plugin layer matters less than the engineering discipline of using them properly. Most WordPress sites are slow because of plugin bloat, not because the wrong cache plugin was installed.
GoGoChimp is a Glasgow-based AI-powered CRO agency that has been running page-speed engineering as part of conversion rate optimisation since 2013. We have shipped page-speed-led engagements for ecommerce (BeeFRIENDLY Skincare, Affordable Golf), B2B (ClickBoost, VectorCloud), and SaaS clients across the UK, Europe, and the US. The Speed Sprint £1,500 engagement is the most common entry point.
Yes. Core Web Vitals (LCP, INP, CLS) have been ranking factors in Google since 2021. The bigger impact is downstream: slow pages lose conversions before search ranking matters. Affordable Golf went from homepage LCP 21.3s to 6.1s in a March 2026 engagement, a 71% improvement. Desktop performance score went 41 to 70. CLS 0.123 to 0.007. Every metric moved into the green band.
Ecommerce page speed has three structural enemies: oversized product images, third-party tag bloat (analytics, chat widgets, recommendation engines, abandoned-cart pop-ups), and bloated theme code. The fix order is image compression and WebP conversion first (highest leverage, lowest risk), then a render-blocking script audit (remove tags nobody is reading), then theme-level code review (Liquid optimisation on Shopify, custom JS deferral). Affordable Golf moved through all three in a phased March 2026 engagement.
Three metrics. LCP (Largest Contentful Paint) measures how long the largest above-the-fold element takes to render; pass below 2.5 seconds. INP (Interaction to Next Paint) measures how responsive the page is to user input; pass below 200ms. CLS (Cumulative Layout Shift) measures visual stability; pass below 0.1. Affordable Golf passed all three after a phased March 2026 engagement (CLS specifically went 0.123 to 0.007).
A CDN is a network of geographically distributed servers that cache and serve static assets (images, CSS, JavaScript) from a location physically closer to the visitor. Round-trip time drops; perceived load speed improves. Cloudflare, Fastly, and AWS CloudFront are the three most-used. The CDN helps least on tiny sites with regional traffic and most on image-heavy ecommerce or video-heavy SaaS marketing sites. ClickBoost compressed video assets from 121.5 MB to 17.2 MB as part of a March 2026 page-speed overhaul; CDN delivery of the compressed assets was the second multiplier.
A 60-second AI scan shows which page-speed issues are leaking conversions on your homepage, and the £/month each one is costing your revenue.
✓ Built on Build Grow Scale's 347-store CRO research
✓ Avg 28-34% lift (expert-led AI CRO benchmark)
✓ Free, no signup
