We Added 9 Analytics Tools to the Same Page. Here Are the Lighthouse Scores.

We wanted to stop arguing about analytics and performance. Every vendor claims their script is “lightweight.” Every marketing page says “no impact on page speed.” So we did what any reasonable team would do: we installed 9 analytics tools on the exact same page and ran Google Lighthouse until the numbers settled.

The results range from invisible to devastating.

Methodology

We started with a baseline HTML page — static, no framework, no images above 50 KB. A clean surface to isolate the variable: the analytics script itself. Then we added each tool's default installation snippet, one at a time, and ran Google Lighthouse 5 times per tool. We took the median of each run.

All tests used Lighthouse's default mobile throttling: a simulated Moto G Power on Slow 4G. This is the same configuration Google uses when evaluating Core Web Vitals for search ranking purposes.

We measured four metrics:

• —Performance Score (0–100) — Lighthouse's composite metric
• —Total Blocking Time (TBT) — milliseconds the main thread is blocked
• —Largest Contentful Paint (LCP) — time to render the largest visible element
• —Interaction to Next Paint (INP) — responsiveness to user input

The results

The gap between the lightest and heaviest tools is 12 points on the Performance Score. Adobe Analytics adds 340 ms of Total Blocking Time. GA4 pushes LCP from 620 ms to 760 ms — a 22.5% increase — from a single script.

What the numbers mean

Google uses Core Web Vitals — LCP, INP, and Cumulative Layout Shift — as ranking signals. A Lighthouse Performance Score below 90 is a yellow flag. Below 50 is a red one. Every point you lose narrows the margin between your site and a competitor's in search results.

The conversion impact is equally direct. Research from Deloitte and Google consistently shows that a 1-second delay in page load reduces conversions by approximately 7%. When an analytics script adds 230 ms of LCP latency (as GA4 does here), that is not a theoretical concern — it is measurable data loss in your revenue pipeline.

For an eCommerce site processing 50,000 sessions per month at a 2.5% conversion rate and a 80 EUR average order value, a 7% conversion drop represents roughly 7,000 EUR in lost monthly revenue. From a measurement script.

Why are enterprise scripts so heavy?

The heaviest scripts in this test share a pattern: they ship features that are either blocked by modern browsers or restricted by privacy regulations. Cross-domain tracking modules. Third-party cookie management. Client-side fingerprinting. Remarketing pixel bundles.

These features exist because they were built in an era when browsers allowed unrestricted tracking across the web. Safari, Firefox, and Brave have since blocked most of this behavior. GDPR and ePrivacy require explicit consent before any of it can execute. But the code is still there, still loading, still blocking the main thread on every single page view — whether or not a visitor has consented to anything.

A cookieless approach eliminates the need for these modules entirely. No cookie management code. No consent-gating logic on the client. No fingerprinting fallbacks. The result is a script that does less on the client because it does not need to.

The compounding problem

These scores reflect the analytics script in isolation. On a real site, the analytics tool is rarely alone. It loads alongside a consent management platform (CookieBot, OneTrust, Didomi), a tag manager (Google Tag Manager, Tealium), and often a remarketing pixel or two.

The combined penalty is not additive — it compounds. A tag manager loads the analytics script, which fires after consent is evaluated, which itself requires a script. In our testing, GA4 loaded through Google Tag Manager with CookieBot consent management pushed the Performance Score from 100 down to 81 — a 19-point drop from the baseline. Adobe Analytics with OneTrust and a single remarketing tag reached 78.

A tool that requires no consent banner, no tag manager intermediary, and no cookie management layer avoids this compounding entirely. The performance score with SealMetrics in that same stack scenario: 100. Because there is no stack.

How we measured

For full reproducibility, here are the exact steps:

1. 1.Created a static HTML page with a single heading, one paragraph of body text, and a 40 KB hero image. No CSS framework, no JavaScript.
2. 2.Deployed to a clean Cloudflare Pages instance with standard caching headers.
3. 3.Added each analytics tool's default installation snippet (from official documentation) to the page's <head> or <body> as instructed by the vendor.
4. 4.Ran Google Lighthouse (Chrome DevTools, version 12.x) 5 times per configuration using default mobile throttling (Moto G Power, Slow 4G).
5. 5.Recorded the median of each metric across the 5 runs.
6. 6.Removed the script, verified the baseline score returned to 100, then added the next tool.

We encourage you to replicate this test. The methodology is straightforward and the tools are free. If your numbers differ significantly, it likely reflects version differences in the analytics scripts — which change frequently. That variability is itself a data point.

The bottom line

The tool you use to measure your site is itself changing what you measure. A 12-point Lighthouse drop is not a technicality — it affects search rankings, Core Web Vitals assessments, and the conversion rate that determines whether your marketing spend produces returns.

SealMetrics captures 100% of traffic through cookieless first-party collection with zero measurable impact on page performance. No consent banner required. No tag manager dependency. The data is complete and the Lighthouse score stays where it was before you added analytics.

If you are running GA4 or Adobe Analytics, you can calculate how much data you are losing to consent rates and script blocking right now. The performance cost is one more variable to factor in — and unlike data accuracy, this one shows up in every Lighthouse report.