Do Lighthouse and PageSpeed Insights really have no impact on rankings?

What’s the difference between a measurement tool and a ranking factor?

Google explicitly distinguishes between diagnostic tools (Lighthouse, PageSpeed Insights, Chrome UX Report) and actual ranking signals. Lighthouse runs in a controlled lab environment, with simulated network conditions and standardized configuration. Its scores reflect a theoretical modeling of the experience, not the actual aggregated data from millions of users.

PageSpeed Insights combines lab data (Lighthouse) and field data (CrUX). But again, the interface and its recommendations serve to diagnose issues, not to directly feed into the ranking algorithm. What matters for ranking are the Core Web Vitals measured on your real visitors, aggregated over 28 rolling days, with the 75th percentile as the threshold.

Why is there so much confusion around these tools?

The confusion arises from the fact that the metrics modeled by Lighthouse (LCP, INP, CLS) are closely linked to the Core Web Vitals, which have been ranking factors since May 2021. Practitioners see a red score in Lighthouse and panic, whereas their site might very well be green on the field data from CrUX.

Google doesn’t make things easier: Lighthouse is constantly evolving — its scoring formulas, weights, and metrics change several times a year. A score of 85 in January may drop to 72 in March without you having touched a line of code. It's a perpetually moving tool, which makes it unreliable as a direct proxy for ranking.

Do these tools then have a concrete utility for SEO?

Absolutely, but they should be used for what they are: debugging and auditing tools, not ranking oracles. Lighthouse excels at identifying critical errors (blocking JavaScript, unoptimized resources, oversized images) that, while they may not directly break your ranking, degrade the actual experience of your visitors.

PageSpeed Insights provides access to CrUX data about your domain — and that data does count for ranking. But look at the field data, not the lab score. If your CrUX is green and your Lighthouse is red, you have a diagnostic problem, not a ranking issue. The reverse (CrUX red, Lighthouse green) is much more concerning.

• Lighthouse and PageSpeed Insights are not ranking factors — they are measurement and diagnostic tools.
• Core Web Vitals measured on your real users (via CrUX) are confirmed ranking signals.
• A Lighthouse score can fluctuate without you making any changes, simply because the tool evolves.
• Use these tools to identify optimization opportunities, not to predict your position in the SERPs.
• Field data (CrUX) always takes precedence over lab data (Lighthouse).

Is this statement consistent with what we observe in the field?

Yes, and it's verifiable. We regularly see sites with catastrophic Lighthouse scores (30-40/100) ranking on the first page for competitive queries because their field Core Web Vitals are good and their other signals (content, backlinks, EAT) are solid. Conversely, sites with 95+ on Lighthouse might stagnate on page 3 if their authority is low.

The nuance — and this is where many go wrong — is that correlation is not causation. If you optimize for Lighthouse and your ranking improves, it’s not the Lighthouse score that caused the progression, but the actual optimizations (loading time, interactivity) you implemented that improved your field Core Web Vitals. Lighthouse was a thermometer, not the remedy.

What gray areas does Google leave unclear?

Google never explicitly states how Core Web Vitals are weighted in the overall algorithm. We know they're a signal among hundreds, and that their weight is relatively modest compared to content relevance or authority. But no numbers, no ranges — just a vague "it's important."

Another gray area: the exact threshold for eligibility for the