Google's Core Web Vitals
Your website is probably losing customers right now,
and you don't even know it.
Google has been quietly scoring your site for years... measuring how real visitors experience it on real devices, over real connections. It's a rolling scorecard based on what your actual customers go through every time they visit, and more than half of all websites fail. Most site owners have no idea because everything looks fine on their office Wi-Fi.
These scores aren't hidden away somewhere... anyone can check yours in about 10 seconds, including your competitors.
Written by Valentin Bora — 25 years building for the web. Trusted with €15M+ in annual revenue and 12M+ monthly visitors across client sites. 5.0/5 on Codeable (only 2% of developers get in, 4.8/5 on Trustpilot).
What Google measures
It comes down to three things.
Loading speed
How fast the main content of your page appears. The big headline, the hero image, the product photo... whatever the largest visible element is. Google wants this under 2.5 seconds. If your customers are staring at a blank screen for 4-5 seconds on their phone... that's a fail.
Responsiveness
How quickly your site reacts when someone taps a button, clicks a link, or types in a form. If there's a noticeable lag between "I tapped Add to Cart" and something actually happening... Google measures that. And not just the first interaction, every single one across the entire visit.
Visual stability
Whether things jump around while the page loads. You've experienced it... you're about to tap a link and suddenly an ad loads above it, pushes everything down, and you tap the wrong thing. Google measures exactly how much your layout shifts and it counts against you. Sites running ads get hit especially hard here... if the ad placements aren't set up properly, every ad that loads late causes a layout shift that Google records.
Google collects all of this from real Chrome users rather than simulations, so a perfect score on a lab test doesn't mean you're passing. Real customers on mid-range phones over cellular connections experience something very different from a simulated test on a fast server. Google uses the real-world data for rankings, not the lab results.
Why this exists
Google genuinely wants a faster web.
Google launched Core Web Vitals in 2020 and started factoring them into search rankings in mid-2021. The reasoning is straightforward... it doesn't matter how fast Google Search is if the pages people find are slow to load, frustrating to use, or visually broken. Your customers deserve a good experience, and Google is incentivizing site owners to deliver one.
They've been refining the metrics since then. In March 2024 they replaced one of the original three because developers had found ways to game it... 93% of sites were technically "passing" while real visitors were still experiencing sluggish interactions. The new version closed those loopholes and the pass rate dropped to 65% overnight. Google just wants the scores to reflect what your customers actually experience.
This has been running for six years now, it's getting more precise over time, and it's built into the infrastructure of how Google evaluates every website. It's not a trend.
What it costs you
Every second of delay
is money walking out the door.
Deloitte studied 30 million user sessions across 37 major brands. A 0.1-second improvement... just a tenth of a second... produced an 8.4% increase in conversions and a 9.2% higher average order value. Think about what 8% more sales means on your current numbers.
Vodafone ran an A/B test with two visually identical landing pages... one optimized for Google's Core Web Vitals, one left as-is. The faster page generated 8% more sales and 15% better lead conversion, with 11% more items ending up in carts.
When your site takes 3 seconds to load instead of 1, 32% more visitors leave before seeing anything. At 5 seconds that jumps to 90%. And pages loading in 1 second convert at 3x the rate of pages that take 5 seconds. Once you're past the 3-second mark, the drop-off gets steep fast. Every page load and every interaction is a small barrier your customers have to clear... they'll keep going if the barriers are low or their motivation is high, but most visitors aren't on a mission, and a slow site is all the excuse they need to leave.
more conversions per 0.1s faster
more bounces at 5s vs 1s load
higher conversion on fast pages
of shoppers leave over slow checkout
The compounding effect
A slow site trains Google
to show your competitors instead.
When someone searches on Google, taps your link, waits too long, gives up, and goes back to tap on a competitor instead... Google tracks that. Their internal system categorizes it as a "bad click" against your site and a "good click" for whoever they went to next.
This isn't speculation. In May 2024, about 2,500 pages of Google's internal ranking documentation were accidentally leaked. It confirmed a system called NavBoost that adjusts your rankings based on how visitors behave after clicking your search result. If visitors bounce back to Google quickly it counts against you, but if they stick around... that works in your favor. Google also confirmed using Chrome browser data as a ranking input... something they'd denied publicly for years.
So a slow website creates a compounding problem. Poor performance produces bad click signals that lower your rankings and reduce your traffic, and at some point it gets hard to justify investing in a fix. The longer you wait, the deeper the hole gets... and your competitors who already fixed this are picking up the traffic you're losing.
WordPress and WooCommerce
If your site runs on WordPress,
the odds are stacked against you.
WordPress powers over 40% of the web, but only 43% of WordPress sites pass Google's Core Web Vitals on mobile... compared to 85% for Duda, 74% for Wix, and 65% for Shopify. The median mobile performance score for a WordPress site is 38 out of 100.
WordPress itself is not is bad software, but the ecosystem stacks up performance problems over time. Every plugin adds code that runs on every page whether it's needed or not. Page builders add megabytes of overhead. And without proper caching configuration, every single page request runs through the full server pipeline from scratch... database queries, PHP processing, the works.
WooCommerce makes it trickier. Cart and checkout pages... the exact pages where your customers are about to buy... can't be cached because they have to be personalized. Cart contents, shipping calculations, tax rules, payment processing... all of that runs in real time. So the pages that matter most for revenue are the ones where caching can't help.
None of this is your fault. Most agencies and developers build sites that look great but never check real-world performance with actual visitors on actual devices. By the time someone looks at the scores, the problems have been compounding for years.
What a fix looks like
It's usually just something loose under the hood... a couple of screws that need tightening.
Audit
I start with a free high-level audit... I pull your scores from Google's data, identify what's failing and give you the big picture of where the problems are. If you want the full breakdown with specific recommendations your developer or webmaster can act on, that's a paid detailed audit you can take to whoever maintains your site.
Fix
Usually it's a combination of hosting, caching, image optimization, and cleaning up plugin bloat. The specifics depend on your site... sometimes hosting is the main bottleneck, sometimes it's the WordPress setup, sometimes it's both. The actual user experience improves right away, but Google's data takes anywhere from 10-14 days up to 28 days to fully reflect the changes.
Monitor
Google updates scores over a 28-day rolling window, so improvements show up gradually. I track progress and make sure the numbers keep moving in the right direction... and that nothing slips back when someone installs a new plugin or changes a page template.
The thing is... speed optimization plugins alone don't solve this. They help with loading speed, but they can't fix responsiveness... that depends on how much code your site forces the browser to run, and no caching plugin changes that. A site that loads fast can still feel sluggish when someone tries to interact with it. Fixing Google's Core Web Vitals properly means looking at the full picture, not just installing a plugin and hoping for the best.
Frequently asked questions
Things people usually ask me about this.
How do I check my site's Core Web Vitals scores?
Go to pagespeed.web.dev and type in your URL. The "field data" section at the top shows your real scores from Google's data... that's what actually matters for rankings. The "lab data" section below is a simulation and can look quite different from what real visitors experience. If there's no field data for your URL, it means your site doesn't have enough Chrome traffic for Google to have collected data yet, but you can still check the lab results as a rough guide.
Do Google's Core Web Vitals actually affect my search rankings?
Yes, but probably not in the way you'd expect. The direct ranking weight of Core Web Vitals is relatively small... content quality and relevance still matter far more. But there's an indirect effect that's much bigger. When your site is slow and visitors bounce back to Google to click on a competitor, Google's NavBoost system registers that as a negative signal against your site. So the ranking damage from a slow site comes mostly from user behavior, not from the score itself.
Why do WordPress sites fail more often than other platforms?
It's the plugin model. WordPress is designed to be extended with plugins, and most business sites run 20-30 of them. Each plugin adds its own code that loads on every page whether it's needed there or not. Stack that on top of a page builder like Elementor or Divi, and you end up with a lot of code competing for the browser's attention. Platforms like Shopify and Wix control the entire stack, so they can optimize things WordPress simply can't.
Will installing a caching plugin fix my scores?
It'll help with loading speed, but it won't fix everything. Caching plugins are great at reducing server response time... they serve a pre-built version of your page instead of rebuilding it from scratch for every visitor. But they can't fix responsiveness issues because those depend on how much JavaScript your browser has to process after the page loads. And if you're running WooCommerce, your cart and checkout pages bypass the cache entirely because they need to be personalized. Caching is part of the solution, but it's rarely the whole thing.
How long does it take to see improvements after fixing things?
Your visitors will feel the difference right away... pages load faster, buttons respond quicker, things stop jumping around. But Google's scores take longer to catch up because they're based on a 28-day rolling window of real user data. Depending on how much traffic your site gets, you'll typically see Google's numbers start moving within 10-14 days, with the full picture settling in around 28 days as the old data cycles out.
Does this affect mobile and desktop differently?
Yes, and mobile is where most sites struggle. Only 48% of websites pass Core Web Vitals on mobile versus 56% on desktop. Mobile devices have slower processors, less memory, and often run on cellular connections... so everything takes longer. Google uses mobile-first indexing, which means your mobile scores are what matter for rankings even when someone searches from a desktop. And for most sites, 60-70% of traffic comes from mobile, so that's where the revenue impact hits hardest.
My developer says our PageSpeed score is 90+. Are we fine?
Maybe, maybe not. PageSpeed Insights shows two different sets of results. The lab score (Lighthouse) is a simulation run under controlled conditions... it can look great even when real visitors are having a poor experience. What matters for Google rankings is the field data at the top of the report, which comes from actual Chrome users visiting your site over the past 28 days. I've seen sites with a Lighthouse score of 90+ that still fail Core Web Vitals in the field because real-world conditions on phones and slow connections are very different from a lab test.
How much does fixing Core Web Vitals typically cost?
It depends on how deep the problems go. I offer a free high-level audit that shows you what's failing and why... think of it as a health check that tells you whether you have a problem and how serious it is. If you want a detailed audit with specific, actionable recommendations you can hand off to your developer or webmaster, that's a paid engagement. The detailed audit is portable... it's yours to share with whoever maintains your site, whether that's me or someone else.
What if my site runs ads? Does that make things worse?
Yes, significantly. Ads are one of the biggest causes of layout shift problems because they often load after the rest of the page is already visible, pushing content around as they appear. If your ad placements aren't reserving the right amount of space before the ad loads, every late-loading ad counts as a layout shift that Google measures against you. News sites, media sites, and content-heavy WordPress sites running ad networks tend to get hit the hardest. The good news is that most ad-related CWV problems can be fixed by properly sizing ad containers and controlling when and how ads load... it's more about configuration than removing ads entirely.
What if my site isn't built on WordPress?
Google's Core Web Vitals apply to every website regardless of platform. The principles are the same... fast server response, optimized images, minimal JavaScript, stable layouts. WordPress just happens to fail more often because of how its plugin ecosystem works. If your site runs on Shopify, Squarespace, or a custom platform, the measurement is identical and the same business impact applies... the fixes just look different depending on what you're working with.
Can I just rebuild my site from scratch instead?
You can, but in my experience it's rarely necessary and usually not worth the disruption. A rebuild takes months, costs significantly more, and introduces all new risks... you're essentially starting from zero with an unproven setup. Most performance problems can be fixed on the existing site with targeted changes. I've taken sites from PageSpeed scores in the 20s up to the 80s and 90s without rebuilding anything... just cleaning up what was already there.
About the author
Valentin Bora
I've been building for the web for 25 years. I've contributed to WordPress Core... the software behind over 40% of the web... and I've been managing performance, infrastructure, and Core Web Vitals for sites handling €15M+ in combined annual revenue and 12M+ monthly visitors, including a news group with 3M+ monthly readers and complex ad infrastructure.
I work through Codeable, an exclusive freelancer network where only 2% of applicants are accepted. Codeable holds a 4.8/5 on Trustpilot from 932+ reviews... for comparison, Upwork sits at 3.8/5. My personal rating across 166 projects on the platform is 5.0/5.
The free high-level audit costs you nothing and takes a few minutes of your time. If you need a detailed audit with specific recommendations, that's a paid engagement... but it's portable. You can take it to your own developer, your agency, or whoever maintains your site. I'm not trying to lock you in.
"I've worked with many developers and engineers throughout my career — Valentin is amazing. I could sense his talent, knowledge, and experience immediately; which is typical of extremely bright developers yet also very rare."
Mike C."Above & beyond what was required. Not just capable but reliable and most of all, an absolute genuine pleasure to work with — of all the developers I've worked with, this is an absolute 1st!"
Kiran B."This guy is a lifesaver! My business was crippled for almost three weeks. Once Valentin and I connected he had my problem solved in a few hours of work."
Tara N.All reviews from Codeable
Your scores are public.
I can pull them right now.
Send me your URL and I'll do a free high-level check... what's passing, what's failing, and whether it's worth digging deeper. If everything looks fine, I'll tell you that too, with a virtual pat on the back!
Or look up your own scores at pagespeed.web.dev
Further reading (sources & references)
- Deloitte — "Milliseconds Make Millions" (2020). 30 million user sessions across 37 brands. Commissioned by Google, published on web.dev.
- Vodafone — Web Vitals A/B test case study (2021). Two identical landing pages, one optimized for Core Web Vitals. Published on web.dev.
- Google/SOASTA — Mobile page load time and bounce probability. 11 million mobile landing pages analyzed.
- Portent — Page load time and conversion rate analysis. 100 million pageviews. Found 4.42% conversion drop per additional second of load time.
- Yottaa — 2025 Web Performance Index. 500 million visits across 1,300+ e-commerce sites. LCP and INP improvements correlated with conversion increases.
- Akamai — Online Retail Performance Report (2017). 10 billion visits analyzed. 100ms delay reduced conversions by 7%.
- Rakuten 24 — Core Web Vitals case study. 53% increase in revenue per visitor, 33% higher conversion rate after CWV optimization. Published on web.dev.
- redBus — INP optimization case study. 7% more sales from a 72% INP improvement. Published on web.dev.
- Lazada — Performance case study. 17% conversion increase from 3x LCP improvement. Published on web.dev.
- The Economic Times — Core Web Vitals case study. 43% bounce rate reduction from CLS and LCP improvements. Published on web.dev.
- Google Algorithm Documentation Leak (May 2024) — ~2,500 pages of internal API documentation. Analyzed by Rand Fishkin (SparkToro) and Mike King (iPullRank). Confirmed NavBoost click signals and Chrome data usage in rankings.
- John Mueller (Google) — Public statements on CWV ranking weight. "It is a ranking factor, and it's more than a tie-breaker, but it also doesn't replace relevance."
- Chrome UX Report — CrUX Technology Report. WordPress CWV pass rates and platform comparisons. Updated monthly by Google.
- HTTP Archive — State of the Web. Overall CWV pass rates across 18+ million websites. Data from July 2025 CrUX dataset.
- Google — Page experience documentation. Core Web Vitals as ranking signals, mobile-first indexing, and user experience guidelines. Updated December 2025.
- Advanced Web Ranking — Ranking correlation study. 3 million+ pages analyzed. Higher-ranking pages consistently had lower LCP values.
- First Page Sage — Q1 2025 algorithm weight research. Estimated page speed at approximately 1% of total ranking algorithm weight.
- Baymard Institute — Cart abandonment statistics. 22% of online shoppers abandon carts due to slow page performance.
- SEMrush — E-commerce CWV assessment data. 60% of e-commerce websites fail Core Web Vitals.
- Statista (2025) — Mobile commerce statistics. Mobile drives 60-78% of e-commerce traffic, representing $4 trillion and 59% of global e-commerce sales.
- Greg Linden (former Amazon engineer) — Internal A/B test estimate (2006). "Every 100ms costs 1% in revenue." No detailed methodology published, but directionally validated by subsequent studies.
- Walmart — Page speed and conversion study (2012). 1-second improvement correlated with 2% conversion increase.
- DebugHawk — Q4 2025 Real User Monitoring report. 5.7 million pageviews. Median PHP execution time of 483ms on WordPress, page caching delivering 7x TTFB improvement.
- SEOPress — Performance case study. Achieved 96% mobile PageSpeed score without a caching plugin. Hosting accounts for ~70% of performance impact.