What is a good PageSpeed score?
Your Google PageSpeed score is one of the most misunderstood numbers in digital marketing. Here is what the test actually measures, why a "perfect 100" is neither achievable nor desirable for a real business website, and which numbers actually affect your revenue and rankings.
For a professionally optimized business site that still runs advertising, analytics, and reviews: mobile Performance of 70–85 and desktop of 85–95 is excellent. Accessibility and SEO should reach 95–100 on any platform. A perfect 100 usually means the site stripped out its revenue tooling to get there — and rankings are driven by real-visitor field data, not this lab number anyway.
What the test actually simulates
The mobile PageSpeed test does not measure your customers’ experience. It simulates a budget Android phone from ~2019, on a slow 4G connection, with the processor artificially slowed down 4×. It is a deliberate stress test — the digital equivalent of testing a car’s highway performance while towing a trailer uphill.
Your actual customers on modern phones and normal connections typically load the same page in 2–3 seconds. The lab test may show 6+ seconds for the identical page.
Scores fluctuate — a single run means very little
PageSpeed runs on Google’s shared servers. The same website, tested minutes apart with zero changes, routinely swings 15–30 points depending on server load and cache state. We have documented identical code scoring 63 and 92 within the hour.
Our measurement standard: run the test three times back-to-back and use the middle score. Any vendor or report citing a single run — high or low — is citing noise.
Why a working business site cannot score 100
Every point below 100 has an owner. On a modern commerce site, most remaining deductions come from software you need and cannot rewrite:
| What the report flags | Who owns it | Why it stays |
|---|---|---|
| "Unused JavaScript" (largest item) | Google Analytics, Google Ads, Meta Pixel | This is your ad tracking. Removing it means your campaigns can’t measure conversions. |
| "Best Practices" deductions, console errors | The platform itself (Shopify, Wix, etc.) | Injected by the platform’s own commerce and analytics infrastructure. Not editable by anyone. |
| Third-party widgets | Reviews, chat, email popups | Each is a business feature traded for a few lab points. |
| "Use next-gen formats (AVIF)" | Platform image CDN | Most platforms don’t offer AVIF yet. |
This is why virtually every Shopify store on the internet — including the largest brands — scores 70–80 on Best Practices, and why mobile Performance in the 70s–80s represents a fully optimized store that still runs advertising.
What "good" actually looks like
Realistic ceilings for a professionally optimized site that runs ads, reviews, and analytics:
| Metric | Optimized target | Notes |
|---|---|---|
| Desktop Performance | 85–95 | Real bandwidth; reflects the code quality |
| Mobile Performance | 70–85 | Capped by required tracking scripts on the throttled test |
| Accessibility | 95–100 | Fully controllable — we take this to the top |
| Best Practices | 70–80 on Shopify/Wix | Platform floor; 90+ only on custom-built sites |
| SEO | 95–100 | Fully controllable — we take this to the top |
| Cumulative Layout Shift (CLS) | 0 | The stability metric users feel most — fully controllable |
The numbers that actually matter
Google Ads quality scores and Google Search rankings are driven by Core Web Vitals field data — measurements collected from your real visitors’ real devices — not by the lab score you see when you type your URL into the test. That field data appears at the top of the PageSpeed page ("Discover what your real users are experiencing") once the site has sufficient traffic.
A site with lab-mobile 75, CLS of 0, and 2-second real-device loads will outperform — in rankings, in ad costs, and in sales — a stripped-down site that scores 95 but removed its conversion tracking to get there.
What professional optimization delivers
When we optimize a site, we take every point that is actually available: right-sized images, prioritized above-the-fold loading, eliminated render-blocking waste, zero layout shift, removal of abandoned apps and duplicate tracking tags, full accessibility and SEO scores. What remains in the report afterward is the fixed cost of the platform and the marketing stack — the same cost every competitor pays.
The goal is not a vanity 100. The goal is the fastest site your business model allows — with your revenue tools intact.
Frequently asked questions
Rarely, and usually not desirable. Most remaining deductions on an optimized commerce site come from required revenue tools: analytics, ad pixels, review widgets, and the platform’s own infrastructure. A 100 usually means conversion tracking was stripped — trading measurable revenue for a vanity number.
PageSpeed runs on Google’s shared servers; server load and cache state swing identical code 15–30 points between runs. Run it three times back-to-back and use the middle score.
No — rankings and Ads quality use Core Web Vitals field data from your real visitors’ devices. The lab score is a diagnostic, not the ranking input.
Mobile Performance 70–85 and Best Practices 70–80 represent a fully optimized store with its apps and tracking intact — that’s where virtually every major Shopify brand lands. Accessibility and SEO should still hit 95–100.