Three caching plugins later, your WordPress site is still pulling 4.2 seconds to first paint — here is what is actually wrong
Most WordPress speed problems are not caching problems. They are page-builder bloat, plugin sprawl, untuned database queries, and font-loading strategies that haven't been touched since 2019. The optimisation that actually moves PageSpeed from 35 to 85 is a structural pass, not a plugin install.
BOOK A SPEED AUDIT CALL12,000+ WordPress sites of speed-optimisation practice Average lift: PageSpeed 35 → 80+ in 4-6 weeks CWV-passing field data, not just lab data
What an honest speed audit covers
The boring measurement layer first
CrUX field data (real Chrome users, not Lighthouse simulations). Core Web Vitals — LCP, INP, CLS — for the top 10 highest-traffic pages. Lighthouse mobile + desktop. Network waterfall analysis. Most agencies start with a Lighthouse score and recommend a caching plugin; the real audit measures what users actually experience and rebuilds against that.
The structural diagnoses that matter
Page-builder weight (Elementor, Divi, WPBakery typically add 200-500KB per page). Plugin sprawl (30+ plugins all loading on every page). Heavy fonts loaded as render-blocking resources. Hero images served at 2x the needed dimensions. Untuned MySQL queries against tables without proper indexes. Third-party scripts (analytics, chat, marketing) blocking the main thread.
The fixes ranked by ROI
Output is a prioritised punch-list with each item scored on impact (PageSpeed-points-gained estimate) and cost (engineering hours). High-impact + low-cost first. Image optimisation, font-display: swap, lazy-loading, deferred third-party scripts. Then mid-impact + mid-cost: replacing the heaviest plugin, tuning the database, configuring object cache. Then high-impact + high-cost: replacing the page builder, replacing the theme, headless rebuild.
When this is not the right call
If your WordPress site is fundamentally fighting the platform — page-builder-driven content modelling, 50+ plugins, custom-coded theme from 2018 — the optimisation engagement is patching a leaky boat. The real answer is the rebuild-or-replatform conversation. The 30-min call surfaces which conversation you should be having.
Frequently asked questions
How much does WordPress speed optimization cost?
A focused one-off optimisation engagement (audit + fix the top 10 issues): 1,800-5,500 USD, ships in 2-3 weeks. A deeper rebuild-style optimisation (rebuilding the theme, replacing 5+ heavy plugins, ground-up CDN + caching strategy): 8,000-25,000 USD over 4-8 weeks. Recurring monthly speed maintenance: typically bundled into mid-tier care plans at 500-1,500 USD/month.
My PageSpeed score is 35 — what is realistic to get to?
Most WordPress sites starting at 30-45 mobile can realistically reach 75-90 mobile within 4-6 weeks of focused work. Above 90 requires structural changes — replacing the page builder (Elementor / Divi / WPBakery) with a native theme, ditching multi-purpose themes, removing 5+ plugins. Achievable but priced accordingly.
Why is my WordPress site slow even on a fast host?
Three usual causes. (1) Page builder bloat — Elementor, Divi, WPBakery all add 200-500KB of JS+CSS to every page. (2) Plugin sprawl — 30+ plugins each loading on every page, often firing the same database queries 3 times. (3) No real caching strategy — page caching alone is not enough; you need object cache (Redis), opcode cache (proper PHP-FPM), and a CDN with proper edge rules. Fast host without these is a fast empty path; the slow part is the application layer.
Should I switch hosts to fix speed?
Sometimes. If you are on shared hosting at 4 USD/month, yes — move to managed WordPress hosting (Kinsta, WP Engine, Pressable, Cloudways at the higher end). If you are already on managed hosting and the site is still slow, the host is not the bottleneck — the WordPress application layer is. Audit before migrating.
When you are ready
Bring your URL + current PageSpeed score (mobile and desktop) + the Search Console "Core Web Vitals" report screenshot. By the end of 30 minutes you will know whether optimisation is the right shape, what the realistic lift is, and the price range.