Wondering if your phone visitors get a slow, jumpy page? Core web vitals mobile comes down to three checks: LCP at 2.5 seconds or less, INP at 200 milliseconds or less, and CLS at 0.1 or less.
This guide sits within the wider topic of mobile website development and focuses on how phones are scored.
| Item | Value |
|---|---|
| LCP, “good” threshold | 2.5 seconds or less (2500 ms) |
| INP, “good” threshold | 200 ms or less |
| CLS, “good” threshold | 0.1 or less |
| Percentile used to judge a page | 75th percentile of page views |
| INP replaced FID | March 12, 2024 |
| Field data window | 28 days |
The thresholds come from web.dev’s threshold definitions (last updated May 7, 2025), and the INP date and data window from the LoadFocus glossary. These figures can change, so check the sources before you rely on them.
What the three metrics measure on a phone
According to web.dev, the Core Web Vitals are three metrics, and each one covers a different part of the experience. On a phone, you feel each of them directly.
- Largest Contentful Paint (LCP) measures perceived load speed. It marks the point when the page’s main content has likely loaded.
- Interaction to Next Paint (INP) measures responsiveness. It captures what users feel when they try to interact with the page.
- Cumulative Layout Shift (CLS) measures visual stability. It quantifies unexpected layout shift of visible content.
INP looks at every click, tap and key press during a visit and reports the longest one, ignoring outliers, as the LoadFocus glossary explains. It replaced First Input Delay (FID) on March 12, 2024, because FID only measured the delay before the first interaction was handled.
CLS reports the largest burst of shifts during the page’s life. So one late-loading banner that pushes the article down can fail a page, as the same glossary notes.
Thresholds and how a page passes
Each metric has a “good” and a “poor” boundary, with “needs improvement” in between. The values below are from web.dev.
- LCP: good at 2500 ms or less, poor above 4000 ms.
- INP: good at 200 ms or less, poor above 500 ms.
- CLS: good at 0.1 or less, poor above 0.25.
A page is not judged on a single visit. web.dev uses the 75th percentile of all page views to that page or site: if at least 75 percent of page views meet the “good” threshold, the metric counts as good. For example, a 75th percentile LCP of 2 seconds is classified as good, while one of 5 seconds is classified as poor.
The LoadFocus glossary adds that a page passes only when all three metrics are good at that percentile. One failing metric is enough to fail the page.
Why mobile results often differ from desktop
The same glossary says each metric is judged separately for mobile and desktop, on what real visitors experienced rather than on one test run. Real visitors arrive on slower phones, weaker networks and from other countries.
That is why a page can feel quick on your laptop and still fail on phones. Your own device is rarely the typical one, so the 75th percentile of real mobile visits is the number to watch.

How to measure core web vitals on mobile
The numbers that count are field data from real Chrome users, collected in the Chrome UX Report (CrUX) over a trailing 28-day window, per the LoadFocus glossary. Lab data comes from a tool such as Lighthouse that loads the page once on a simulated device and network, so it helps you find causes rather than judge the result.
- Check field data in PageSpeed Insights, which pulls from CrUX, as the Website Spec advises.
- Open the Core Web Vitals report in Google Search Console, which shows the same kind of field data.
- Run a lab test to see what is slow and why.
- Use the web-vitals JavaScript library to send real-user metrics to your analytics.
Pages without enough Chrome traffic return no field data. And because the field data is a 28-day window, a fix shows up gradually, so judge a change by lab results first.
Fixes that help mobile scores
Most fixes fall under one of the three metrics. The short lists below are drawn from the LoadFocus glossary and the Website Spec.
Improve LCP
- Make the LCP image discoverable in the initial HTML, and preload it if it is only referenced from CSS or JavaScript.
- Never lazy-load the LCP image. It must load eagerly.
- Shrink the file, and cut render-blocking CSS and synchronous scripts in the head.
Improve INP
Long tasks on the main thread are the usual cause of poor INP, often from script loading and parsing. Do as little work as possible in event handlers, update the screen first and defer the rest, and keep the DOM small.
Test beyond the homepage too. The Website Spec notes that interaction-heavy pages such as forms and search results are usually the worst offenders.
Improve CLS
The usual causes are images and videos without dimensions, fonts that render at a different size than their fallback, ads or widgets that resize themselves, and content inserted above what the visitor is reading. The fix is to set width and height (or an aspect ratio) and reserve space for anything that loads late.
For wider load-time techniques, see our guide to mobile page speed optimization. Layout choices also shape stability, and mobile website design best practices looks at them from the design side.
Do Core Web Vitals affect mobile rankings?
As of the Ahrefs guide updated January 5, 2025, mobile page experience and the included Core Web Vital metrics have officially been used for ranking pages since May 2021. That makes them a real part of mobile search, not only a usability nicety.
Still, treat them as one input among many. The LoadFocus glossary points out that Google’s documentation does not say how much weight they carry. Good scores will not rescue content that fails to answer the query.
Frequently asked questions
What is a good LCP on mobile?
Show answer
An LCP of 2.5 seconds or less, measured at the 75th percentile of real visits, according to the LoadFocus glossary. Above 4000 ms is classed as poor by web.dev.
Did INP replace FID?
Show answer
Yes. The LoadFocus glossary states that INP replaced First Input Delay as a Core Web Vital on March 12, 2024.
Why does my page pass a lab test but fail in the field?
Show answer
A lab test loads the page once on a simulated device and network. Real visitors arrive on slower phones and weaker networks, so field data can look worse, as the LoadFocus glossary explains.
How long until a fix shows in my report?
Show answer
Field data covers a trailing 28-day window, so a fix shows up gradually. Use lab results to check a change sooner, as the LoadFocus glossary suggests.



