Every 100ms costs you conversion. We give them back.
A 1-second delay in mobile load time reduces conversion by up to 20%. A PageSpeed score below 50 is a direct SEO ranking penalty. Most sites scoring in the 20s and 30s have the same set of problems: unoptimised images, render-blocking JavaScript, no caching strategy, a bloated theme, and a hosting environment that was never designed for performance. We've seen these patterns across hundreds of sites and know exactly where to look, what to fix, and what order to fix it in. Most clients see 40–60 point PageSpeed improvements within the first engagement.
What's included
- Core Web Vitals audit (LCP, CLS, INP)
- Image optimisation & next-gen format delivery
- JavaScript bundle analysis & reduction
- Server-side caching & CDN configuration
- Font loading strategy
- Third-party script audit & deferral
How we deliver
- 1Full performance audit & baseline report
- 2Prioritised fix list with impact estimates
- 3Image pipeline & optimisation implementation
- 4Caching, CDN & hosting configuration
- 5JS/CSS audit & bundle reduction
- 6Post-optimisation report & monitoring setup
Technologies we use
- Next.js
- Webpack Bundle Analyzer
- Lighthouse
- WebPageTest
- Cloudflare
- Vercel
- imgix
- Sharp
- Partytown
- Fontaine
Why Origin for Website Speed Optimization
Ranked fix list with impact estimates
We don't give you a generic Lighthouse report — we give you a prioritised list of fixes ranked by expected PageSpeed impact. You know exactly what each change will deliver before we start.
We fix the causes, not just the symptoms
Caching plugins mask slow pages; they don't fix them. We trace every performance problem to its root cause and fix it there — the result is a faster site, not just a higher score on one tool.
Real-device testing, not just lab simulations
Lab scores don't represent what your users experience. We test on real mobile devices on real connections and deliver field data alongside lab data in every audit.
Industries we serve
“We were scoring 22 on mobile PageSpeed and could see in the analytics that users were bouncing. Origin spent four weeks on optimisation — we're now at 81 on mobile, 96 on desktop. Organic traffic is up 28% in three months.”
Frequently asked questions
- Our PageSpeed score is 28 — what's actually causing it?
- For most sites scoring below 40, the top culprits in order of impact are: unoptimised images (wrong format, wrong size, no lazy loading), render-blocking scripts (third-party analytics, chat widgets, fonts loaded synchronously), excessive JavaScript shipped to the browser, and no server-side caching (every page request hits the origin). We audit all of these in the first week and produce a ranked fix list.
- Can you improve performance on our existing WordPress / Shopify site, or do we need to rebuild?
- Usually significant improvement without a rebuild. For WordPress, the standard playbook includes: proper caching plugin configuration (WP Rocket or W3 Total Cache), image format conversion to WebP, plugin audit (removing what's unused), CDN setup, and PHP version upgrade. For Shopify, the toolkit is more limited but image optimisation, script deferral, and theme code review usually move the needle significantly.
- What are Core Web Vitals and why do they matter for SEO?
- Core Web Vitals are the three speed metrics Google uses as a ranking signal: Largest Contentful Paint (LCP — how fast the main content loads), Cumulative Layout Shift (CLS — how stable the page is while loading), and Interaction to Next Paint (INP — how quickly the page responds to user input). Sites that pass all three thresholds get a ranking advantage. Sites that fail all three are at a ranking disadvantage against competitors with better scores.
- We pass Core Web Vitals in the lab but fail in the field — what does that mean?
- Lab data (Lighthouse, PageSpeed Insights) runs in a controlled environment with a simulated connection. Field data (Chrome User Experience Report) is from real users on real devices and connections — often much slower mobile connections than the lab simulates. Field failures usually indicate the lab test is too optimistic about what real users experience. We test with actual mobile devices and throttled connections to understand real-world performance.
- How do you handle third-party scripts that are slowing down the site but we can't remove?
- Defer loading until after the page is interactive, where possible. For scripts that genuinely need to load earlier (payment processors, auth), we analyse their impact and optimise loading order to minimise their effect on LCP and INP. We use Partytown for analytics and tag manager scripts — this moves them off the main thread so they don't compete with page rendering.