Core Web Vitals started as Google’s attempt to replace a messy pile of speed scores with a small set of user-centred signals. The practical side for marketers: these three field metrics now sit between your landing page, your search visibility, your ad delivery and your conversion rate.
The three metrics that count
As of September 2026, the core set is LCP, INP and CLS. Each has a clear job and a firm threshold.
- Largest Contentful Paint (LCP): loading. Good is 2.5 seconds or less; poor starts above 4 seconds.
- Interaction to Next Paint (INP): responsiveness. Good is 200 milliseconds or less; poor starts above 500 milliseconds.
- Cumulative Layout Shift (CLS): visual stability. Good is 0.1 or less; poor starts above 0.25.
These are not lab scores. Google evaluates them from Chrome User Experience Report field data, using opted-in Chrome sessions and a 28-day rolling window. A page only passes when at least 75% of page views clear the good threshold for all three metrics. That is why a fix deployed today can take weeks to show up fully.
Why lab scores can lie to you
PageSpeed Insights shows a Lighthouse lab estimate alongside real-user CrUX data, and the two routinely disagree. One 2025 client case described by Noel Ceta saw the lab score jump from 65 to 92 while LCP worsened from 1.8 to 4.2 seconds. The culprit was lazy loading applied to the hero image, followed by a 20% traffic decline. It is a useful warning: a green lab score is not the same as a fast page for actual visitors.
Ads and Core Web Vitals
Ad operations can quietly damage all three metrics. Ads are requested asynchronously and inserted after content has rendered, which makes them a common source of CLS. Google’s publisher guidance says the most effective fix is to reserve space for the largest creative that could serve, even if a smaller creative leaves blank space. Client-side header bidding also runs auctions in the browser, adding load and responsiveness pressure that feeds into LCP and INP.
Google AdSense made the revenue link explicit in a March 24, 2026 email, pushing back/forward cache, the Speculation Rules API and AI-assisted debugging to publishers. The email cited a 27% rise in page views at Netzwelt and 9% mobile revenue growth at Yahoo! JAPAN News.
What this means for your growth team
Shopify’s April 27, 2026 analysis across its merchant base found conversion was about 3.5% lower for every additional 100 milliseconds of LCP. Stores loading in 2.5 seconds converted roughly 30% below stores at 1.5 seconds, though the slowest 5% of stores were excluded and the figures are aggregate platform data, not a controlled experiment.
Paid media campaigns inherit the same pages. A slow landing page can lose visitors before an analytics tag fires, distorting conversion data and the bidding models trained on it. Behaviour analytics tools now allow session filtering by LCP, INP and CLS, so performance teams can use these metrics next to engagement and revenue data.
The ranking weight itself is small. Google’s Martin Splitt has described the ranking effect as very small in most cases, and Google’s documentation says there is no single page experience signal. Still, for high-intent e-commerce and lead-gen pages, the user-level impact is often much larger than the ranking effect.
A simpler playbook
Stop optimising only for the Lighthouse number. Start with CrUX field data in Search Console or the CrUX API. Prioritise the hero content path for LCP, remove aggressive lazy loading above the fold, reserve ad space to limit layout shift, and cut long JavaScript tasks for INP. Then wait a full 28-day window before judging the result. Treat Core Web Vitals as revenue guardrails, not just an SEO audit item.
Source: PPC Land
