Click2.ai® Free audit

Topic · Performance 2026

Core Web Vitals fix.
The honest playbook.

What LCP, INP, and CLS actually measure. The thresholds Google uses. The five causes that fail most sites. The fixes that actually move the numbers — and the structural answer when the tactical fixes don't.

Free Core Web Vitals audit See the thresholds ›
Core Web Vitals fix LCP INP CLS

What CWV is

Three metrics. Real users. 75th percentile.

Core Web Vitals is Google's measurement of how a real page feels to a real visitor. Three metrics: how fast the main content paints (LCP), how responsive the page feels to clicks and taps (INP), and how stable the layout stays (CLS). Google scores each at the 75th percentile of your real-user traffic — meaning 75% of your visitors must hit the "good" threshold for that page to pass.

  • 2.5sLCP good threshold
  • 200msINP good threshold
  • 0.1CLS good threshold

Why it matters

2026 reality
"Failing Core Web Vitals doesn't just hurt SEO. It cuts AI Overviews inclusion and tanks paid-ad quality scores. One slow site, three problems."
Click2.aiPerformance practice

Thresholds

Good · needs work · poor.

Metric What it measures Good Needs work Poor
LCPLargest Contentful Paint< 2.5s2.5–4.0s> 4.0s
INPInteraction to Next Paint< 200ms200–500ms> 500ms
CLSCumulative Layout Shift< 0.10.1–0.25> 0.25

Scored at the 75th percentile of real user traffic in Chrome (CrUX). Lab tools (PageSpeed Insights, Lighthouse) estimate; the field data is what Google ranks on.

1 · LCP

Largest Contentful Paint fixes.

Fix 1

Edge CDN for the hero image

Origin servers in one region kill global LCP. Edge CDN serves the hero from the location nearest the visitor.

Fix 2

AVIF / WebP with width & height

Modern formats are 30–60% smaller than JPEG. Explicit dimensions let the browser allocate space immediately.

Fix 3

Preload the hero

<link rel="preload" as="image" href="..."> tells the browser to start the download before parsing the rest of the HTML.

Fix 4

Eliminate render-blocking above the fold

Defer non-critical CSS. Async-load non-critical JS. Inline the critical CSS for the hero.

Fix 5 (structural)

Move off shared / managed WordPress hosting

A static site on a global edge CDN delivers sub-1s LCP by default. Most of the tactical work above becomes unnecessary.

Diagnostic

PSI + CrUX + WebPageTest

PSI for the quick view, CrUX for real-user field data, WebPageTest for waterfall debugging.

2 · INP

Interaction to Next Paint fixes.

Cause #1

Third-party JavaScript

Chat widgets, analytics, A/B testing, tag managers, heatmaps — they steal the main thread. Audit, remove what's not earning its weight, async-load the rest.

Fix 1

Break up long tasks

Any task > 50ms blocks input. Split into smaller pieces, yield with scheduler.yield() or setTimeout(0).

Fix 2

Move work off main thread

Web workers for heavy computation. requestIdleCallback for non-critical work that can wait until the browser is idle.

Fix 3

Debounce input handlers

Heavy work on every keystroke is INP poison. Debounce or throttle to fire once per N ms.

Fix 4 (structural)

Ship less JavaScript

A WordPress site with 8 marketing plugins ships 1–3MB of JS. A static site usually ships under 100KB. No tactical fix beats shipping less.

Diagnostic

Chrome DevTools Performance + RUM

DevTools for synthetic profiling. Real User Monitoring (web-vitals library) for the field data that actually counts.

3 · CLS

Cumulative Layout Shift fixes.

Fix 1

Width & height on every media element

Images, videos, iframes, embeds — all must declare dimensions so the browser reserves space before they load.

Fix 2

Reserve space for ads

min-height on ad containers. Don't let ad insertions push content around.

Fix 3

No content inserted above existing content

Banners, cookie bars, A/B test variants — all reserve space above the fold rather than pushing content down.

Fix 4

Font loading strategy

font-display: optional or preload + size-adjust to prevent FOUT-driven layout shift.

Fix 5

Use transform, not top/left

Animations on transform and opacity don't trigger layout. Animations on top, left, width, height do.

Diagnostic

PSI + DevTools rendering panel

PSI flags the largest CLS culprits. DevTools rendering panel visualizes shifts in real time.

FAQ

Straight answers about Core Web Vitals.

What are the thresholds in 2026?

LCP < 2.5s good, INP < 200ms good, CLS < 0.1 good. Scored at the 75th percentile of real user traffic.

What is INP and why did it replace FID?

INP measures input-to-next-paint latency across the entire visit, not just the first interaction. Replaced FID in March 2024 because FID didn't measure visual response time.

Fastest LCP fix?

Edge CDN + AVIF/WebP hero + preload + defer render-blocking. Structural: move off shared WP hosting.

Fastest INP fix?

Cut third-party JS. Break up long tasks. Move heavy work off the main thread. Structural: ship less JavaScript.

Fastest CLS fix?

Width/height on every media element. Reserve space for ads and injected content. font-display strategy.

Do CWV affect ranking?

Yes — confirmed Page Experience signal. Tiebreaker among pages of similar relevance. Also affects AI Overviews inclusion and paid-ad quality scores.

Can WordPress pass?

Yes but takes aggressive optimization. Static / JAMstack passes by default. For most small businesses, migration is faster and cheaper than the WP optimization project.

Free audit

Get your real CWV numbers + a fix list.

Send your URL. We pull CrUX field data, run lab tests, and ship a prioritized fix list — tactical and structural.

Free site audit Or: cost calculator ›