ADBLOAT vs Google PageSpeed Insights

PageSpeed Insights grades a URL from the publisher's perspective — how fast the page loads, what Core Web Vitals look like, whether render-blocking scripts slow the initial paint. ADBLOAT grades from the visitor's perspective — how much ad-tech weight is still reaching the browser after the visitor's defenses. The tools complement each other. Confusing them is how a site can score 95 on PageSpeed and still feel bloated to real users.

Summary table

PageSpeed Insights is a publisher tool

PageSpeed Insights reports on a URL — how that URL performs for the median visitor. Google serves it via Lighthouse (synthetic lab audit) plus Chrome User Experience Report (CrUX) field data. The output grades Core Web Vitals (Largest Contentful Paint, Interaction to Next Paint, Cumulative Layout Shift) plus a list of opportunities and diagnostics. Publishers use it to prioritize engineering work: cut render-blocking JS, lazy-load images, defer third-party scripts.

Crucially, PageSpeed Insights tests the URL in a default browser environment. It does not know — or care — what defenses the visitor uses. The score reflects the worst-case performance experienced by the user with no ad blocker, no VPN, no fingerprint protection. A site can score 90 on PageSpeed (good Core Web Vitals) and still load 3MB of tracker payload that gets blocked by half the visitors' extensions, making real-world performance much better than the PageSpeed score suggests.

ADBLOAT is a visitor tool

ADBLOAT flips the question. Instead of grading a URL, it grades a defense. The fixture loads 46 known tracker endpoints — the same endpoints a real ad-tech-heavy page would load — in a sandboxed iframe, using the visitor's actual browser with their actual extensions, DNS filters, and VPN active. Every blocked request is counted toward the grade; every request that completes is counted against it. The composite: 60% bytes blocked + 25% handshakes prevented + 15% payload reduced, compared to a naked-Chromium baseline.

ADBLOAT explicitly excludes load time from its composite. The reason: load time depends on the visitor's network speed, not on their defense. A visitor on fiber who blocks 90% of ad-tech gets roughly the same ADBLOAT grade as a visitor on LTE who blocks 90% — even though their load times are wildly different. PageSpeed's grade would call them different; ADBLOAT calls them the same because their defense is the same.

When to use PageSpeed Insights

• You publish a site and want it faster. PageSpeed's diagnostics list exactly which scripts to defer, which images to resize, which third-parties to lazy-load. It is the canonical publisher tool.
• You care about Google ranking. Core Web Vitals are a confirmed Google ranking factor since 2021. Publishers who ignore PageSpeed lose organic traffic.
• You're auditing a competitor. PageSpeed runs on any public URL, so you can benchmark competing sites without any cooperation from them.
• You're optimizing for mobile users specifically. PageSpeed's mobile profile (throttled CPU + slow 4G) reveals which sites actually work on phones in realistic conditions.

When to use ADBLOAT

• You want to know how much ad-tech your defenses still let through. Every blocker claims "blocks 99% of ads." ADBLOAT measures the actual number against a reference workload.
• You're comparing two blockers or two DNS filters. Run ADBLOAT with each setup. The grade delta is the honest comparison.
• You're justifying the time investment in a layered stack. Pi-Hole alone vs Pi-Hole + uBlock vs Pi-Hole + uBlock + router-level DoH blocking. ADBLOAT quantifies each layer's contribution.
• You suspect a VPN-bundled ad blocker is theater. Many consumer VPNs advertise "ad blocking" as a feature with minimal actual effect. ADBLOAT tells you whether their claim is real.

Using both is the thorough approach

A publisher who cares about the user experience should run PageSpeed Insights on their own pages AND have their QA team run ADBLOAT with common ad blockers active. This catches two different failure modes: PageSpeed flags structural performance issues; ADBLOAT flags whether the site's ad-tech integrations are so dense that even well-defended users see degraded behavior.

A privacy-minded visitor should use ADBLOAT to tune their own defenses over time, and occasionally look at PageSpeed Insights results for sites they visit often to understand why some pages feel heavy — sometimes the answer is bad site engineering (PageSpeed's domain), sometimes it's ad-tech density (ADBLOAT's domain).

FAQ

Try both

Related reading: The Real Cost of the "Free" Internet · Ad-Tech Glossary · ADBLOAT vs WebPageTest