Does Google Analytics Slow Down Your Site? We Measured It — Here's What Happened

The Measurement

Every site with Google Analytics installed is running the exact same trade: real visitor data, in exchange for page speed. Most site owners never see the bill. We did — because we watched our own score move.

The Impact, In Numbers

Before we touched anything, our homepage scored:

• 90/100 on Mobile PageSpeed
• 95/100 on Desktop PageSpeed

Lighthouse (the tool behind Google’s PageSpeed Insights) flagged one specific culprit: 67 KiB of unused JavaScript, all coming from the Google Analytics (GA4) tracking script. After we changed when that script loads — not whether it loads, just when — both scores hit a clean 100/100.

Ten points on mobile. Five on desktop. Same site, same content, same design. The only thing that moved was the timing of one script tag.

What Your Site Is Actually Getting From That Tracking Connection

GA4 exists to answer a simple question: what are people doing on your site? Where they came from, what they clicked, how long they stayed, whether they converted. That data shapes real decisions — what pages to invest in, what’s not working, which marketing channels are worth the spend.

The problem is how that data gets collected. The GA4 script has to load early enough to catch a visitor’s first actions, so by default it loads immediately, right alongside your actual content. On a fast connection you’d never notice. On a phone over 4G, that script is competing with your hero image and your headline for the same limited bandwidth — and Lighthouse counts every byte of it that doesn’t get used in those first critical seconds as dead weight against your score.

That’s the trade every analytics-enabled site is making, whether the owner knows it or not: better data, worse first impression.

Why This Matters Beyond the Number

A PageSpeed score isn’t a vanity metric. It’s a stand-in for what a real visitor experiences in the first couple of seconds on your site — and increasingly, it’s also a signal search engines and AI answer engines use to decide whether your site is worth surfacing at all. A slow-loading page doesn’t just lose a visitor; on a bad connection, it can lose them before your content even finishes painting.

So the instinct to “just remove analytics” is understandable, but it’s the wrong fix. You lose the data that tells you whether the site is actually working.

The Better Fix: Load On Intent, Not On Arrival

Instead of loading GA4 the instant a page opens, we changed the trigger. Now the script only loads once a visitor does something — scrolls, clicks, or moves their mouse. We paired that with a short fallback timer, so visitors who land and bounce without interacting still get counted, just a few seconds later than before.

The logic is simple: if nothing is competing for load time until a visitor shows real intent, the initial render is faster — and analytics still captures everyone who actually engages with the page.

This connects to something we’ve written about before: how to prove user interest to Google. The core idea is the same one at work here. What search engines and AI systems are really trying to measure isn’t just whether a page loaded — it’s whether a real person found it worth engaging with. A scroll, a click, a few extra seconds on the page: those are the signals that separate a genuine visit from a bounce. Delaying analytics until one of those signals fires doesn’t just protect your PageSpeed score — it aligns your tracking with the exact moment interest becomes measurable.

The Takeaway

If your site runs Google Analytics — and most do — it’s worth checking whether it’s costing you PageSpeed points the same way it was costing us. The fix isn’t dropping your data. It’s being deliberate about when you collect it.

We log every performance change we make to this site, wins and setbacks both, on our Performance Log — including this one, with links to the confirmed 100/100 results.

Want someone to actually watch these numbers for your site? Apply to Work With Us →