What do Web Vitals measure?
Web Vitals help describe loading, interaction responsiveness and visual stability. The Core Web Vitals are LCP, INP and CLS. INP replaced FID in March 2024 and considers eligible interactions throughout a visit rather than only the first input delay.
How performance relates to search rankings
Google’s page experience guidance
Google’s 2021 Page Experience announcement provides historical context. Core Web Vitals are used by Google’s ranking systems, but good values do not guarantee a leading position. Relevant, useful content and other factors still matter.
Slow loading or shifting content can make a task harder. Measure bounce rate, engagement and conversions to understand your own audience, using the definitions of your analytics system. Do not treat these measurements as proof of a direct ranking effect.
Longer time on a page is not automatically better: a user may complete a task quickly or remain because the interface is confusing. Social interactions and repeat visits should not be presented as automatic ranking signals on the basis of this article.
Benefits to verify
A faster, stable interface can reduce friction and help users complete tasks. Whether it changes trust, conversions or repeat visits must be measured on your site. A small performance gain is not a guaranteed competitive or search advantage.
Treat Web Vitals as diagnostic evidence. An improvement in a metric is not, by itself, evidence of increased organic traffic.
LCP, INP and CLS
LCP — Largest Contentful Paint: when the largest eligible visible content element is rendered. A good value is 2.5 seconds or less.
INP — Interaction to Next Paint: responsiveness to clicks, taps and keyboard input, through the next paint. A good value is 200 milliseconds or less. It does not measure the full completion time of a later asynchronous request.
CLS — Cumulative Layout Shift: a measure of unexpected layout shifts, not a mechanism that prevents them. A good value is 0.1 or less; this is a unitless score.
For real-user assessment, use the 75th percentile of visits, separating mobile and desktop. These three metrics describe particular aspects of experience, not every part of performance or functionality.
Separate laboratory and field data
A laboratory test reproduces selected conditions. Field data describes actual visits and may cover a different period and set of devices. Check each value’s source, URL and aggregation scope before comparing results. Missing data is not a passing result.
An automated interaction test covers the actions it executed. It does not establish the INP of all visitors. Keep untested flows and missing field measurements visible in the result.
Choose a measurable correction
- Resources: investigate oversized images and expensive scripts or styles. Preserve necessary functions and do not lazy-load the likely LCP image.
- Delivery: examine server response time, cache behavior and network distance. Use caching and CDNs where appropriate; private responses require suitable access and cache policies.
- Layout: test representative mobile and desktop layouts and reserve space for images or inserted content.
- Checks: combine repeatable lab runs with available field data. Recheck important pages and interactions after changes.
Verify after publication
Repeat the same URL under comparable conditions and confirm that the task still works. Record the publication date and other site changes. Evaluate search traffic separately, using comparable periods; a technical improvement cannot guarantee more visits.
Start with one relevant page and one plausible cause. Use Milten’s audit findings to guide the investigation, then verify the published change. For current definitions and limits, see Google’s page experience guidance and the Web Vitals guide.