Core Web Vitals in Plain Language - What to Fix First

LCP, INP, and CLS explained without jargon, plus a sensible order of attack for real business sites and how to measure what users actually feel.

21 March 2026 , Blackgate

Core Web Vitals are Google’s label for a small set of user-centred metrics: loading, responsiveness, and visual stability. They matter for experience and they align with what search systems reward when everything else is equal - but the main reason to care is that slow, jumpy pages lose people before they read your offer.

Today’s set is LCP, INP, and CLS. (Older articles may still say FID for interactivity; INP replaced FID as a Core Web Vital - if your tool shows FID, upgrade your reporting to INP where possible.)

The three metrics, in plain terms

LCP (Largest Contentful Paint) - How long until the main content of the page is visibly loaded (usually a hero image, video poster, or large text block). If LCP drags, people stare at a blank or half-empty screen.

INP (Interaction to Next Paint) - How snappy the page feels when someone taps, clicks, or types - how long until the next visual update after that input. Heavy JavaScript, main-thread blocking, and huge third-party scripts hurt this.

CLS (Cumulative Layout Shift) - How much stuff jumps around while the page loads (images without dimensions, late-loading ads, fonts swapping layout). Bad CLS means mis-taps and cheap-feeling UX.

Where to read them (field data first)

  • Google Search Console - Core Web Vitals report uses Chrome real-user data (CrUX) for URLs that get enough traffic. Start here for “what real visitors see.”
  • PageSpeed Insights - Shows lab scores plus field data when available. Good for URL-level checks and suggestions.
  • Lighthouse (DevTools) - Lab test on your machine; useful for debugging, not a substitute for field data.

RUM (real user monitoring) from your analytics or APM tool adds value if you need segments (device, country, logged-in users).

What to fix first (practical order)

There is no universal rule, but on typical marketing and content sites the highest impact sequence is usually:

  1. LCP - Slow server or CDN, huge unoptimised hero images, blocking render, fonts delaying text. Fix the main template first (home, top landers), then roll patterns out.

  2. INP - Too much JavaScript on the main thread, long tasks, non-deferred third parties, bulky client frameworks doing work on every interaction. Trim, defer, code-split, and measure again.

  3. CLS - Reserve space for images and embeds, load fonts without layout thrash, avoid injecting banners or cookie bars that push content mid-read.

If Search Console flags one metric badly on your money URLs, start there - data beats a generic checklist.

Tactics that move the needle

LCP: Compress and right-size images, use modern formats where supported, cache HTML at the edge, improve TTFB (hosting, database, application work), avoid hero carousels that delay the real LCP element.

INP: Audit third-party tags, break up long tasks, defer non-critical scripts, ensure interactions are not blocked by unnecessary work on the main thread.

CLS: Width/height on images and video, stable slots for ads and embeds, careful font loading (e.g. font-display and matching fallbacks).

Ongoing: Performance regresses when marketing adds tags and plugins without a budget. Treat new scripts like production code - who owns them, and what’s the performance cost?

Contact us if you want help prioritising fixes, tuning hosting and caching, or cleaning up tag weight without breaking measurement.

Have a Project for Us?

Tell us where you're headed, and we'll show you how we can help you get there.

Let's chat