Core Web Vitals affect rankings indirectly – mostly through how users behave on your site, not through the raw LCP or FCP numbers themselves. You can test this yourself with a simple caching experiment, and the results will show you exactly how much speed is worth to your traffic.
What Are Core Web Vitals?
Core Web Vitals are a set of Google metrics that measure how fast a page loads and how stable it feels to a user, including Largest Contentful Paint (LCP) and First Contentful Paint (FCP). Google uses these signals as part of its ranking system, but there's ongoing debate about exactly how much weight they carry.
PageSpeed documentation in details
Three Theories on How Core Web Vitals Influence Rankings
SEO practitioners generally point to three possible mechanisms behind how landing page optimization drives traffic gains:
- Direct metric impact - the raw LCP and FCP numbers themselves factor into ranking calculations.
- Behavioral signals - Google observes how users interact with your site through field data (PageSpeed Insights field data reflects real user sessions), and that behavior shapes rankings.
- Bounce-driven impact - slow-loading pages cause users to leave before the page finishes loading, and that abandonment itself hurts rankings.
How to Test This With a Simple Caching Experiment
You don't need to guess which theory is correct. Run this experiment on your own site:
- Enable caching for real users only. Serve bots the uncached version and give users a faster, cached version.
- Measure the traffic change from Google. Expect a modest increase – this is roughly the gain attributable to improved user behavior alone.
- Enable caching for everyone, including bots. A CDN like Cloudflare works well for this.
- Measure again. This time, expect a more substantial traffic increase.
The gap between these two results is the answer: page speed for bots matters just as much as page speed for users, and often more than SEOs assume.
Why Bot-Facing Speed Matters Too
If your crawl-facing pages load slowly - even while your user-facing pages are fast - that gap can cost you rankings. This also cuts the other way: moving from a fast host to a CDN that actually slows your site down can hurt your positions, not help them.
Takeaway: keep your load times competitive with, at minimum, the top three sites in your target keyword cluster – for both users and bots.
Why You Should Check Individual Pages, Not Just Site Averages
A site-wide average can hide serious problems. Your homepage might load quickly while a key landing page, loaded down with heavy scripts, video, or unoptimized images, drags – and the average won't catch it.
Check your most important pages individually using two data sources in PageSpeed Insights:
- Field data – real, historical performance data from actual users.
- Lab data – a snapshot of performance in the current moment.
Then compare both data sets against your top competitors for the keyword you're targeting. That comparison shows you exactly where you're falling behind.
What to Do With the Data
PageSpeed Insights will point you toward specific fixes once you know where you stand. If the recommendations aren't clear, two options work well: bring in a specialist, or use a well-structured AI prompt to get a detailed breakdown of what each recommendation means and how to act on it.