Google's metrics sound technical, but they describe something simple: how quickly the visitor sees something useful, how quickly the page responds, and how stable it stays while it loads.
Core Web Vitals are three measurements Google collects from real visitors using Chrome. They do not come from a test at your office; they come from your customers, on their phones and their networks. That is why they matter: they describe the experience anyone who visits your store actually has.
The three thresholds at a glance
- LCP (loading): good under 2.5 seconds.
- INP (responsiveness): good under 200 milliseconds.
- CLS (stability): good under 0.1.
A page passes when at least 75% of visits are within the thresholds. So it is not enough for it to be fast for you on the office Wi-Fi; it has to be fast for the customer with an old phone on the bus.
LCP, how quickly the main content appears
Target: under 2.5 seconds on mobile. In most stores, the LCP element is the main image or the product title. The usual causes of delay: an image without loading priority, fonts that block rendering, and a slider that loads three photos before showing the first.
The fixes that pay off first: the main image in a modern format (WebP or AVIF) at the right size, with high loading priority and no lazy loading; fewer CSS and JavaScript files that must download before anything appears; and a server that responds quickly, with caching for pages that are the same for everyone.
INP, how quickly it responds to a tap
It replaced the old FID and is stricter: it measures every interaction, not just the first. If "Add to cart" takes half a second to react, it sees it. The usual culprit: too much JavaScript from tracking scripts.
Chat, pop-ups, review widgets and ad pixels all run on the same thread as the button the customer taps. The more of them load together, the more likely the button has to wait. The answer is not to remove them all; it is to load them later, only where needed, through a tag manager such as Google Tag Manager, with consent.
CLS, how much the page jumps
The most irritating for the user and the easiest to fix. Give images and ad slots explicit dimensions, and reserve space for banners that appear later.
A classic example: the customer is about to tap "Buy", the cookie banner or a promo bar appears above, the page shifts and they tap something else. Fonts that change size once they load have the same effect on a smaller scale.
Where to see them
- Google Search Console, in the Core Web Vitals report: it groups your pages into "good", "needs improvement" and "poor", with real user data.
- PageSpeed Insights: at the top it shows the page's real user data (when there is enough) and below it the lab test with suggestions.
- Lighthouse in Chrome: useful for testing a change before it goes live.
What to ignore
The overall PageSpeed score is a blend of lab measurements and changes from load to load. Useful as a compass, useless as a target. Rankings are influenced by real user data, not by the number you saw once on your laptop.
Also ignore suggestions that promise a few milliseconds at a high cost. If the three metrics are green for your real visitors, the next hour of work pays off more in copy, product photos and checkout than in one more point on the score.
The order to fix them in
- CLS first, because it usually takes a few lines: dimensions on images and fixed space for banners.
- LCP next: the main image, fonts, fewer render-blocking files.
- INP last, because it needs a review of third-party scripts and often a conversation about which ones are really needed.
In practice, most causes of delay in a store are the same three: plugins, images and the database. We go through them in why your store is slow. If you want us to take it on, see ecommerce development and a real example in Elixir Makeup.
Frequently asked questions
Do Core Web Vitals affect Google rankings?
Yes, as part of the page experience assessment, alongside many other factors. A fast site will not outrank more relevant content, but between comparable pages it counts.
Why does Search Console say "not enough data"?
Real user data only appears when a page or site has enough Chrome traffic. On smaller sites use the lab test as a guide, on mobile and a slow network.
How quickly does an improvement show?
Google's data is an average of the last 28 days, so a fix shows fully after about four weeks. The lab test shows it immediately.
Is INP the same as FID?
No. INP replaced FID in March 2024 and measures responsiveness across all interactions in a visit, not just the first, which is why it is more demanding.
