● Findings · October 2026 · Daniel the Tuner

Why is my Shopify PageSpeed score so low?

You ran your store through Google PageSpeed, saw a number in the 30s or 40s, and your stomach dropped. Before you hire anyone or install a speed app, it helps to know what that number is measuring. It also helps to know how often it disagrees with what your real shoppers experience.

9.9 s vs 1.9 smain content on a typical store: in Google’s test, and for real shoppers

42%of stores scoring 30–49 still pass all of Google’s real-shopper checks

INPthe real-shopper check stores fail most often is reacting to taps, not loading

The score comes from one simulated visit

The PageSpeed score comes from a single lab test.
Google loads your homepage once on a simulated mid-range phone over a deliberately slow mobile connection, then scores five timings.
It’s a stress test. It’s designed to show problems, not to describe a typical visit.
That’s why almost every real store looks slow in it. Among the stores we tested, the typical mobile score was 56, and only 1.7% reached 90.

Your real shoppers are measured separately

Google also collects how fast your pages load for real Chrome users over the past 28 days. This is the Core Web Vitals section at the top of your PageSpeed report, when your store has enough traffic.
It checks three things:
- how fast the main content shows (LCP, should be under 2.5 s);
- how fast the page reacts to a tap (INP, under 200 ms);
- whether the layout jumps (CLS, under 0.1).
The two can be far apart. Across 4,998 stores with real-shopper data, the main content took a typical 9.9 seconds in the lab test but 1.9 seconds for most real shoppers.

Low score, fine store? Often.

Under 30

23% · 276 stores

30–49

42% · 1,659 stores

50–69

57% · 2,343 stores

70–89

65% · 641 stores

90 or more

60% · 79 stores

Share of stores that pass all three of Google’s real-shopper checks, by mobile PageSpeed score. Stores with Google real-shopper data only.

Two things stand out:

Plenty of low scorers are fine. 42% of stores scoring 30–49 pass every real-shopper check.
A higher score stops helping past a point. Stores scoring 90+ did no better with real shoppers than stores scoring 70–89. Chasing the last points can cost more than it’s worth, and some “guaranteed 90+” tricks raise the number without helping a single shopper.

When a low score does matter

It matters when your real shoppers fail too. Among stores that don’t pass:

slow reactions to taps (INP) is the most common failure: 1,270 stores;
then slow main content (LCP): 943;
then layout jumps (CLS): 480.

That order is useful. Slow taps usually come from too much script running on the page: apps, tracking tags, chat widgets. Slow main content is usually images, video, or the main picture waiting in line behind everything else.

What to check, in order

Open your PageSpeed report and look at the top section first (real-shopper data). If it says your store passes, speed work is a nice-to-have, not an emergency.
If it fails, see which check fails.
- Taps (INP): look at your apps and scripts first.
- Main content (LCP): look at images, videos and what loads before the first screen.
- Layout (CLS): look at banners, pop-ups and images without set sizes.
No real-shopper data yet? Small stores often don’t have enough Chrome visitors for Google to report. Then the lab test is all you have. Use it to find specific problems, not as a grade.
Ignore small swings. The same store can score 10 points apart on back-to-back runs.

Check your store: real-shopper results first, then what slows it down →

How we measured

5,991 live Shopify stores, October 2026. Of those, 4,998 had Google real-shopper data.
One Google PageSpeed mobile run per store.
Real-shopper figures are Google’s own 28-day Chrome data as reported by PageSpeed Insights.
Typical values are medians.