Performance Optimisation
UNIT10 optimises real, measured performance — not a single Lighthouse score chased in isolation, but the metrics that actually affect users and rankings.
Why performance is never really 'done'
Site speed degrades gradually — a new tracking script here, an unoptimised image there, a plugin update that adds weight — until a site that launched fast is noticeably slower a year later without anyone deciding to make it that way. Performance optimisation is diagnostic work: finding what's actually causing slowness on your specific site right now, not applying a generic checklist.
We measure real-world performance (not just a synthetic lab score) because the two can diverge — a site can score well in an isolated test and still feel slow to a real visitor on a real mobile connection.
The biggest, most common wins
Unoptimised images are the single most common cause of slow load times on business websites — full-resolution photos served at display size instead of properly compressed and responsively sized. Beyond that, unnecessary JavaScript, render-blocking resources, and poorly configured hosting or CDN setup account for most of the remaining gap between a fast site and a slow one.
- Modern image formats (WebP/AVIF) and responsive sizing
- Reducing and deferring non-critical JavaScript
- Font loading strategy to prevent layout shift
- Efficient hosting and CDN configuration
- Caching strategy tuned to how the site is actually used
Frequently asked questions
What's a good Lighthouse or PageSpeed score to aim for?
A score of 90+ is a reasonable target for most business sites, but the score itself is a proxy — what actually matters is real Core Web Vitals performance for real visitors, which we measure and optimise directly rather than chasing a lab score in isolation.
How much does slow load time actually cost in lost business?
Industry research consistently shows conversion rate drops sharply as load time increases, particularly on mobile and for paid traffic where every second of delay is directly costing ad spend efficiency. The exact impact varies by business, but slow load time is reliably correlated with lost conversions across virtually every study on the topic.
Can you fix performance without a full rebuild?
Often, yes — image optimisation, script reduction and hosting configuration frequently deliver significant improvement without a rebuild. If the underlying platform or codebase is the fundamental bottleneck, we'll say so honestly rather than applying surface fixes to a structural problem.