Core Web Vitals, against field data
WordPress Speed Optimisation and Core Web Vitals Repair
Get a WordPress site measured against what real visitors experience, not a lab score. We work Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift from Chrome field data, fix the render path, the images, the database and the origin, and hold it there.
Get a free written site audit
Send the address and we return a written report inside one working day: patch backlog, exposure, load timings, restore status.
- Worked against CrUX field data, not a one-off PageSpeed number
- Server response time first — it caps everything downstream
- Before and after figures, from the same source, in writing
- Held under monitoring, because speed regresses with every plugin added
What this covers
- Server response time and hosting configuration
- Full-page and object caching, correctly invalidated
- Render-blocking CSS and JavaScript removed from the critical path
- Images resized, compressed and served in modern formats with correct dimensions
- Database cleaned: transients, revisions, orphaned meta, unindexed queries
- Plugin audit — the slow ones, measured, not guessed
- Font loading and layout shift
- Monitoring so a regression is caught within the month
Start at the origin, not the front end
Most WordPress speed advice starts with images and caching plugins. That is the second half of the job. A server taking 1,200ms to produce the HTML leaves the fastest front end in the world starting 1,200ms late, and every cache miss pays it again.
We measure server response first, and we do not report a fix until it holds on uncached requests.
What actually slows a WordPress site down
- Page builders. On one site we measured, the builder accounted for 76–84% of the HTML — around 120KB of JavaScript in a 150KB document.
- Unbounded plugin JavaScript loaded on every page regardless of whether that page uses it.
- Images at full camera resolution scaled down in the browser.
- A database nobody has touched in five years — post revisions, expired transients, orphaned meta.
- Shared hosting oversold to the point where response time depends on what the neighbours are doing.
Which of those a given site is failing on shows up in field data rather than a lab score — the guide to Core Web Vitals on WordPress covers reading it. Performance work runs monthly on the Professional and Business tiers rather than as a one-off, because a tuned site drifts back. It sits alongside hardening and scanning in the same WordPress maintenance retainer.
How the work runs
- Measure the field, not the labWe start from CrUX data for real visitors on real connections. A lab score on a fast machine hides exactly the problems users have.
- Fix the origin firstServer response time is the floor. Nothing on the front end gets under it, so it is the first thing worked.
- Cut the critical pathRender-blocking CSS and JavaScript, then the images that sit in the largest contentful paint.
- Then the databaseTransients, revisions and unindexed queries, which is where long-lived WordPress sites quietly lose their response time.
- Re-measure from the same sourceBefore and after, from field data, so the improvement is real rather than a better lab run.
- Hold itEvery plugin added is a potential regression. Monitoring catches it in the month it happens rather than the year.
Plans that include this
- Starter$39/monthor $390 a year — two months freeBlogs, portfolios and brochure sites
- UpdatesMonthly
- BackupsWeekly, 30-day retention
- Uptime5-minute checks
- Malware removalNot included
- Most chosenProfessional$79/monthor $790 a year — two months freeSmall businesses and lead-generation sites
- UpdatesWeekly + PHP version management
- BackupsDaily, 60-day retention
- Uptime1-minute checks
- Malware removalIncluded
- Business$149/monthor $1490 a year — two months freeStores, membership sites, anywhere downtime costs money
- UpdatesWeekly, tested on staging first
- BackupsDaily, 90-day, multi-location
- Uptime1-minute + 2-hour restore
- Malware removalIncluded
What clients say
They found malware two of our previous agencies had missed, cleaned it in an afternoon, and told us exactly how it got in.
Operations Director, professional services, 4 sitesThe staging step is the whole point. Three years and not one update has taken our booking form down.
Practice Manager, healthcareWe resell it to 31 clients under our own brand. The reports go out with our logo and our clients have never heard their name.
Agency Founder, digital agency
The service that covers this
wordpress core web vitalsPerform Core Web Vitals repair on WordPress against field data rather than a lab score. Real visitor measurements decide the outcome, and a perfect PageSpeed run…
Related pages
Questions people ask before calling
Why is my PageSpeed score different every time?
Lab scores vary with the machine, the network and the run. Field data — what Chrome records from real visitors — is what search actually uses, and it is far more stable.
Is a perfect score worth chasing?
No. The thresholds that matter are LCP under 2.5s, INP under 200ms and CLS under 0.1 for most visitors. Points above that are vanity.
Will a caching plugin fix it?
It helps and it is not enough. Caching cannot fix a slow origin on an uncached request, oversized images, or a plugin running queries on every page load.
How much faster, realistically?
On a typical unoptimised WordPress site we see server response drop from over a second to under 200ms, and LCP move into the green. Sites already tuned see less.
Does speed actually affect rankings?
It is a ranking signal and a smaller one than relevance. Its bigger effect is on conversion — and on crawl budget, which matters if the site is large.