Why Page Speed Still Matters for Google Rankings

Clients sometimes push back when I bring up page speed early in a project. It feels like a technical detail that can be fixed “later.” I used to let that slide more than I should have, until I actually re-read what Google says about it directly instead of relying on secondhand SEO advice.

The three numbers that matter

Google measures real user experience through three Core Web Vitals metrics:

• Largest Contentful Paint (LCP) — how long the main content takes to load. Good is under 2.5 seconds.
• Interaction to Next Paint (INP) — how quickly the page responds when someone taps or clicks. Good is under 200 milliseconds.
• Cumulative Layout Shift (CLS) — how much the page visually jumps around while loading. Good is under 0.1.

Google confirmed these as an official page experience ranking signal back in 2021, and they’re still very much active in 2026. What changed recently is that the thresholds got stricter, and INP in particular became a more heavily weighted metric than it used to be.

Is it actually a big deal, or just a checkbox?

Here’s the honest, unhyped version: content quality and relevance still matter more than speed for ranking. Core Web Vitals function more like a tiebreaker. But when two pages are otherwise competitive, that tiebreaker is exactly what separates page one from page two, and pages with poor INP have been shown to drop several ranking positions after Google’s stricter 2026 threshold update.

What this looks like in practice

Most speed problems on small business sites come from a small list of repeat offenders: unoptimized images, third-party scripts loaded on every page whether they’re needed or not, and fonts or animations that block the page from becoming interactive quickly. None of these require an engineering team to fix. They require someone to actually check, using Google’s own PageSpeed Insights tool, rather than assuming a site is “fast enough” because it looks fine on a fast office wifi connection.

If your site was built more than a couple of years ago and nobody’s specifically tested it against these three metrics since, it’s worth five minutes to check. It’s usually the cheapest ranking improvement available, precisely because it’s a one-time technical fix rather than an ongoing content or backlink effort.