A faster website is better for visitors. The harder question is whether improving your Google PageSpeed Insights score will help you rank higher.
The honest answer: it can help, especially when your site is genuinely slow or fails Core Web Vitals. The PageSpeed Insights score itself is a diagnostic score, not a Google ranking position.
A page scoring 100 does not automatically outrank a more relevant, more trusted page scoring 70. Search results still depend heavily on content quality, search intent, crawlability, internal links, backlinks, and authority. Speed supports that work. It cannot carry a weak page by itself.
This guide explains what PageSpeed Insights measures, which parts matter for SEO, and how to decide whether speed work belongs at the top of your list.
What PageSpeed Insights measures
Google PageSpeed Insights tests a single page and reports on performance, accessibility, best practices, and SEO. Most people focus on the performance score because it is the big number at the top.
That score runs from 0 to 100:
| Score | Rating |
|---|---|
| 90 to 100 | Good |
| 50 to 89 | Needs improvement |
| 0 to 49 | Poor |
The score is useful, but you need to know where it comes from. PageSpeed Insights combines two kinds of information.
Field data comes from real Chrome users over time, when Google has enough data for that page or origin. This is the closest view of what actual visitors experience across real devices and connections.
Lab data comes from Lighthouse, which runs a simulated test under controlled conditions. Lab data is excellent for debugging because it gives repeatable clues, screenshots, timings, and recommendations.
Field data tells you what happened to real users. Lab data helps you find likely causes.
What Core Web Vitals are
Core Web Vitals are Google’s main performance measurements for user experience. They focus on three things: loading speed, responsiveness, and visual stability.
| Metric | What it measures | Good result |
|---|---|---|
| Largest Contentful Paint (LCP) | How quickly the main content appears | 2.5 seconds or less |
| Interaction to Next Paint (INP) | How quickly the page responds after interaction | 200 milliseconds or less |
| Cumulative Layout Shift (CLS) | How much the layout unexpectedly moves | 0.1 or less |
Google’s Core Web Vitals documentation explains these metrics in more detail. The key point for site owners is that Core Web Vitals are measured at the 75th percentile. In plain English, most visits need to have a good experience, not only your best test run on a fast connection.
That matters because one perfect Lighthouse test does not prove your visitors are having a good experience. A page may test well in the lab but still feel slow to people on older phones, weak mobile connections, or busy pages with ads and scripts.
Does the PageSpeed score affect rankings?
Google does not rank pages by simply sorting Lighthouse scores from highest to lowest.
Google’s page experience guidance says Core Web Vitals are used by its ranking systems, but good scores do not guarantee top rankings. The same guidance also says chasing a perfect score purely for SEO may be a poor use of time once the page already gives visitors a good experience.
That distinction is important. PageSpeed Insights can reveal ranking-related problems, but the displayed score is not the ranking factor by itself.
Think of it like a dashboard warning light. If the light is on, you should investigate. If the light is off, polishing the dashboard will not make the car faster.
Field data matters more than one test score
Near the top of a PageSpeed Insights report, look for the Core Web Vitals assessment. It will usually say the page passed, failed, or does not have enough data.
That assessment matters because it is based on real-world data when available. The Lighthouse score lower in the report is based on one simulated run.
This creates a few common situations:
- A page scores 95 in Lighthouse but fails Core Web Vitals for real visitors.
- A page scores 75 in Lighthouse but passes Core Web Vitals in the field.
- A newer or smaller website has no field data because it does not have enough eligible Chrome traffic.
- A page’s score changes between tests because the server, network, ads, fonts, or third-party scripts behaved differently.
Use Lighthouse to diagnose. Use field data, Search Console, and real-device testing to judge whether visitors are actually getting a good experience.
When speed improvements can help SEO
Speed work is most likely to support rankings when it fixes a real user experience problem.
Your Core Web Vitals are failing
If important pages fail LCP, INP, or CLS, performance work deserves attention. A failing page experience signal can become more important when competing pages are similarly relevant and useful.
For example, if two local plumbers have similar service pages and similar authority, the page that loads faster, responds better on mobile, and does not jump around may have an edge.
The ranking lift is usually modest. The business lift can be much larger because visitors can actually use the site.
The site feels slow on mobile
Many PageSpeed problems show up harder on phones than on desktops. Mobile devices may have slower processors, smaller screens, weaker connections, and less tolerance for heavy JavaScript.
Common mobile performance issues include:
- A large hero image delaying the first useful content.
- Too much JavaScript running before the page becomes usable.
- Fonts loading late and causing text to shift.
- Third-party chat, booking, analytics, or ad scripts blocking the main thread.
- Images loading at desktop size on a small phone screen.
Fixing those issues can improve search performance indirectly because visitors stay longer, read more, submit forms, and trust the site more.
Important content loads too late
Some sites technically load, but the meaningful content appears late. This often happens when a page depends heavily on JavaScript, a slow API call, a video background, or a large image above the fold.
Search engines can render JavaScript, but delayed or fragile rendering still creates risk. Visitors also make fast decisions. If the first screen is blank, unstable, or covered by scripts, they may leave before the page has a chance to help them.
Performance audits can reveal whether your headings, intro text, product images, forms, or calls to action appear quickly enough.
Layout shifts make the page hard to use
Layout shift happens when elements move unexpectedly while the page loads. A visitor tries to tap a button, an image loads above it, and the button jumps.
Common causes include:
- Images without set width and height.
- Ads or embeds loading into spaces that were not reserved.
- Fonts swapping after text has already appeared.
- Banners or pop-ups pushing content down.
Reducing layout shift improves the experience immediately. It also helps with CLS, one of the Core Web Vitals.
When chasing a higher score is a bad trade
A PageSpeed score of 35 often points to serious problems. Moving from 35 to 80 can be valuable because the work usually fixes slow loading, poor mobile usability, and unstable rendering.
Moving from 92 to 100 is different. Once a page passes Core Web Vitals, loads quickly on real devices, and works well for visitors, the last few points may not be worth the cost.
Those final points are often tied to tradeoffs:
- Removing a live chat tool that helps customers ask questions.
- Stripping analytics that the business uses to measure campaigns.
- Replacing an embedded booking system that drives leads.
- Spending development time on tiny savings while important pages still have weak content.
- Delaying content improvements, internal linking, or technical crawl fixes.
Score chasing can feel productive because the number moves. Rankings usually move when the page becomes more useful, easier to access, and easier to trust.
Page speed still needs the SEO basics
A fast page with thin content will still struggle.
Before treating PageSpeed as the main ranking lever, make sure the basics are in place:
- The page answers a real search intent.
- The title tag and meta description are specific.
- The page has one clear topic and useful headings.
- Google can crawl and index it.
- Internal links point to it from relevant pages.
- Canonical tags use the correct URL and your site consistently uses one preferred hostname.
- The XML sitemap includes the preferred version.
- The site works on mobile.
- The business has credible backlinks, reviews, or brand signals.
If those pieces feel unfamiliar, start with what SEO actually involves before obsessing over a performance score. If you want a quick baseline check before deeper SEO work, run the 5-minute website health check.
What to fix first
Start with the fixes that affect real visitors and important pages.
| Priority | What to check | Why it matters |
|---|---|---|
| 1 | Pages that fail Core Web Vitals | These are the clearest page experience issues |
| 2 | Key landing pages and service pages | These pages drive leads, sales, and rankings |
| 3 | Mobile loading and responsiveness | Mobile users often feel performance problems first |
| 4 | Oversized images and video | Large media commonly delays the main content |
| 5 | Slow server response | A slow server delays everything after it |
| 6 | Third-party scripts | Ads, chat, tracking, and embeds can block interaction |
For most websites, the practical target is stable, good performance across real visits. A score in the 80s or 90s with passing Core Web Vitals is usually healthier than a fragile 100 achieved by removing useful features.
Common mistakes
Testing only the homepage. Your homepage may be fast while service pages, product pages, blog posts, or location pages are slow. Test the pages that matter for search and conversions.
Looking only at desktop. Desktop scores often look better because computers are faster and connections are stronger. Mobile results usually reveal the problems visitors feel first.
Treating one Lighthouse run as final. Scores move between tests. Run more than one test and look for patterns, not a single number.
Ignoring field data. If real-user Core Web Vitals fail, do not dismiss the issue because one lab test looked good.
Removing useful business tools for tiny gains. A booking widget that slows the page slightly may still be worth keeping if it produces leads. Improve how it loads before removing it.
Fixing speed while ignoring content. A fast page still needs helpful content, clear structure, crawlability, and authority.
Practical next step: audit three pages
Open PageSpeed Insights and test three URLs:
- Your homepage.
- Your most important service, product, or landing page.
- A blog post or informational page you want to rank.
For each page, write down:
- Whether the Core Web Vitals assessment passes or fails.
- The mobile performance score.
- The LCP, INP, and CLS values.
- The top two recommendations that appear more than once across the three pages.
If the same issue appears repeatedly, start there. For example, if every page flags oversized images, compress images and serve appropriately sized versions. If every page shows slow server response, talk to your hosting provider or developer. If third-party scripts dominate the report, review which scripts still earn their place.
After you make changes, wait for field data to update before judging the full result. Lab tests can improve immediately, but real-user data is collected over time.
FAQ
Should every page score 100 in PageSpeed Insights?
No. A perfect score can be satisfying, but it is rarely the best SEO target. Aim for fast, stable pages that pass Core Web Vitals and work well for real visitors.
Is a score of 70 bad?
It depends on the page and the field data. A score of 70 may be acceptable if Core Web Vitals pass and the page feels fast on real devices. It deserves investigation if users are waiting, tapping unresponsive buttons, or seeing content jump around.
How long does it take for speed improvements to show in Google data?
Lab tests can show changes right away. Field data takes longer because it reflects real user visits over time. Expect to monitor Search Console and PageSpeed Insights across weeks, not minutes.
Can speed fixes improve conversions even if rankings do not move?
Yes. Faster pages can help visitors read, browse, enquire, book, or buy with less friction. That can make performance work worthwhile even when the ranking effect is small.