Google Core Web Vitals: What Actually Moves the Numbers

Core Web Vitals turned page speed from a vague virtue into three measurable numbers. That was useful. It also produced a great deal of anxious optimisation of things that were never the problem.

Here is the short version of what these metrics measure and what actually fixes them.

The three metrics, plainly

**Largest Contentful Paint** — how long until the main thing on the screen appears. On most sites this is the hero image or the headline.

**Interaction to Next Paint** — how quickly the page responds when someone taps or clicks. Poor scores here usually mean too much JavaScript competing for the main thread.

**Cumulative Layout Shift** — how much the page jumps around while loading. Caused by images and embeds without reserved space, and by fonts swapping late.

What fixes them, in rough order of impact

• **Serve images properly.** Correct dimensions, modern formats, explicit width and height attributes, lazy loading below the fold. This one change fixes a large share of both LCP and CLS problems.
• **Cut third-party scripts.** Chat widgets, tag managers, heat maps, five analytics tools. Each one costs main-thread time. Audit them and remove what nobody has looked at in six months.
• **Reserve space for anything that loads late.** Ads, embeds, banners, cookie notices.
• **Handle fonts deliberately.** Preload the one or two faces you need, use a sensible display strategy, and do not load nine weights you never use.
• **Cache and compress at the edge.** A CDN in front of the site is the cheapest performance win available.

Measure the right thing

A lab score from a test tool is a diagnostic, not a grade. What counts is field data — what real visitors on real devices and real networks experience. A site can score well in a lab on a fast connection and badly for the majority of its actual audience on mid-range phones.

Check the field report before and after every change, and give it four weeks to settle.

Perspective

Core Web Vitals are a tiebreaker, not a ranking strategy. A fast page with nothing worth reading will not outrank a slow page that answers the question. The reason to fix them is that slow, jumping pages lose customers — measurably, at the checkout and the enquiry form.

Do the image work, cut the scripts you do not need, and stop there unless the numbers say otherwise. That covers most of it for most sites.