Core Web Vitals 2026: What's Changed and What It Means for Your Rankings

Google’s Core Web Vitals continue to evolve. INP has now fully replaced FID as the interactivity metric, thresholds have tightened in places, and new signals are being tested for future inclusion. If you haven’t checked your B2B site’s performance since 2024, there’s a good chance you’re failing metrics that directly affect your rankings. Here’s the 2026 state of play.

What’s Changed Since 2024

The headline change is that INP is no longer new. It replaced First Input Delay as an official Core Web Vital in March 2024, and by now Google’s tooling, Search Console reporting, and ranking systems treat it as fully established rather than experimental.

The three current Core Web Vitals in 2026:

Metric	Good	Needs Improvement	Poor
LCP (Largest Contentful Paint)	under 2.5s	2.5s–4.0s	over 4.0s
INP (Interaction to Next Paint)	under 200ms	200ms–500ms	over 500ms
CLS (Cumulative Layout Shift)	under 0.1	0.1–0.25	over 0.25

The thresholds themselves haven’t moved dramatically, but the bar in practice has, because more of the web has improved. HTTP Archive’s ongoing tracking of Core Web Vitals pass rates shows steady improvement across the web over the past few years, which means the relative bar for “good enough” keeps rising even where the absolute numbers stay the same. Sites that were adequate in 2024 can now be mid-pack or worse simply because competitors caught up.

Google has also continued testing additional signals around responsiveness and smoothness beyond the three official metrics, though none has been promoted to a ranking factor yet. The safest approach is to treat the current three as the baseline and not wait for the next one to force a rebuild.

INP Deep Dive: The Metric B2B Sites Struggle With Most

INP measures how quickly your page responds to any interaction across the full session, not just the first one. Click a nav item, open a dropdown, submit a form; INP captures the worst of these responses, not just the average.

B2B sites tend to struggle here more than consumer sites for a specific reason: forms, filters, and interactive dashboards are common on B2B marketing and product pages, and every one of those is a chance for the main thread to be busy when a user interacts.

The most common causes:

Third-party scripts running on the main thread. Live chat widgets, analytics tags, heatmap tools, and A/B testing scripts often execute JavaScript that blocks interaction response. A user clicks while one of these scripts is running, and the click visibly lags.

Hydration on JavaScript-framework sites. Server-rendered React or Vue pages have to re-run their full JavaScript bundle client-side to attach event listeners before the page becomes properly interactive. Until hydration finishes, clicks feel unresponsive. This is one reason frameworks that ship minimal or zero JavaScript by default, like Astro, tend to post better INP scores out of the box: there’s no hydration cost to pay.

Long tasks blocking the thread. Any JavaScript task over 50ms delays whatever the user tries to do next. These accumulate quietly as more scripts get added to a site over time.

Fixes: audit and remove third-party scripts you don’t actively use, defer or lazy-load the rest, and break long tasks into smaller chunks so the browser can respond to interactions between them.

LCP in 2026: Updated Best Practice

LCP measures how long the largest visible element, usually a hero image or headline, takes to render. The fundamentals haven’t changed, but the priority order has sharpened.

Image optimisation remains the single highest-impact fix. Serving your hero image in a modern format at the correct dimensions, rather than an oversized JPEG, is still the most common quick win on B2B sites. Our image optimisation guide covers this in detail.

Server response time (TTFB) matters more than it used to, because Google’s field data increasingly weights the early part of the load timeline. Static sites served from a global edge network start with a structural advantage here over dynamically rendered pages that query a database on every request.

Render-blocking resources — stylesheets and synchronous scripts in the <head> — still need to be deferred or minimised. And preload hints for the LCP element itself remain underused: telling the browser to fetch your hero image immediately, rather than discovering it partway through HTML parsing, can shave meaningful time off LCP with a single line of markup.

CLS Updates and Common Causes on B2B Sites

CLS measures visual stability — how much content jumps around as the page loads. The measurement window and calculation have had minor refinements over the years, but the practical causes on B2B sites remain consistent.

Images without declared width and height attributes are still the most common culprit; the browser doesn’t reserve space, so content shifts when the image arrives. Web fonts that swap in late cause text reflow. Cookie consent banners that inject above the fold after initial render push everything down. Embedded videos and iframes that load without known dimensions cause the same problem. All of these are preventable with reserved space and explicit dimensions, and none require complex engineering to fix.

Measurement and Monitoring in 2026

Three tools cover what most B2B teams need.

PageSpeed Insights gives you both a lab score and, for sites with enough traffic, real field data from the Chrome UX Report (CrUX). The field data is what matters for rankings; the lab score is useful for debugging specific issues.

Search Console’s Core Web Vitals report shows your CrUX data over time, grouped by URL, so you can see which pages are failing and track whether fixes are working.

web-vitals.js, Google’s own library, lets you capture real user metrics directly in your own analytics if you want visibility beyond the 28-day CrUX window or need page-level detail CrUX doesn’t surface.

Set a simple routine: check Search Console monthly for regressions, run PageSpeed Insights after any deploy that touches your key pages, and treat a 500ms+ LCP regression or an INP creeping past 300ms as a priority fix, not a backlog item.

Where Most B2B Sites Are Losing Points

In practice, the sites we audit fail Core Web Vitals for a small, repeatable set of reasons: unoptimised hero images, third-party scripts loaded without any deferral strategy, and JavaScript frameworks that ship more code than the page actually needs. None of these require a full rebuild to fix, but they do require someone auditing the actual field data rather than guessing.

Want to know exactly where your site is losing Core Web Vitals points? Book a performance audit and we’ll walk through your real CrUX data with a prioritised fix list.