Speed & Core Web Vitals Retainers

One-off performance audits produce reports that get partially implemented and then regress as new content, new plugins and new features add weight. Sustained retainer relationships catch regressions as they happen and prevent the cumulative degradation single-event audits cannot stop. Brands serious about performance treat it as ongoing engineering discipline rather than periodic project.

Lighthouse and PageSpeed Insights produce lab measurements under idealised conditions. Real users on Nigerian mobile networks experience meaningfully worse performance than lab tests show. We optimise to real-user metrics from Chrome User Experience Report and field-collected data rather than chasing lab scores that do not reflect user experience.

Image optimisation produces some of the highest-impact performance wins. Lazy loading, modern image formats (WebP, AVIF), proper sizing and responsive image attributes together typically lift LCP scores substantially.

Third-party script management often introduces performance issues that internal optimisation alone cannot fix. Marketing tags, analytics, chat widgets and conversion pixels each add load. We audit and rationalise these as part of optimisation work.

Database query optimisation supports backend performance underpinning frontend speed. Slow queries produce slow time-to-first-byte regardless of frontend optimisation. We optimise both layers.

CDN configuration tuning extracts additional performance from already-deployed CDNs. Cache policies, edge locations and origin connections all influence delivery speed.

Modern hosting infrastructure with proper edge caching, HTTP/3 protocol support and modern image format negotiation contributes to baseline performance that frontend optimisation alone cannot achieve. We coordinate infrastructure alongside frontend work.

Mobile-specific optimisations matter substantially. Touch-target sizing, viewport configuration, mobile-specific JavaScript pruning all influence mobile performance scores differently from desktop.

Continuous performance monitoring across the site as opposed to one-time audits prevents regression as new features and content get added. We integrate monitoring into ongoing operations.

Specific Core Web Vitals targets - LCP under 2.5 seconds, INP under 200ms, CLS under 0.1 - translate into specific engineering work that we plan against rather than treating performance as vague aspiration.

Performance budget enforcement through CI/CD pipelines prevents new code from regressing performance metrics. Code that fails performance budgets blocks deployment, enforcing discipline at the source rather than discovering regressions later.

Service worker strategies for repeat visitors produce instant-feeling performance through aggressive caching. We deploy service workers where the visitor pattern supports.

Performance budgets become organisational discipline beyond technical implementation. Marketing teams adding tracking pixels, content teams adding heavy media, design teams adding animation - each must understand their contribution to performance budgets.