You spent a weekend compressing images and deleting apps. You open your Shopify speed report expecting a reward, and the score moved three points. Or didn't move at all. It's one of the most deflating moments in running a store, and it sends a lot of merchants down a rabbit hole of chasing a number that was never the right goal.
Before you assume your work failed, you need to understand what that score actually is, because most of the frustration comes from believing it measures something it doesn't.
What the Shopify speed score actually measures
The number in your admin comes from Google Lighthouse, the same engine behind PageSpeed Insights. Shopify runs it against a few sample pages, typically your home page, a product page, and a collection page, then combines them into one score weighted by how much traffic each gets.
The key word is lab. Lighthouse loads your store in a simulated environment with a fixed pretend device and a fixed pretend network. It is a controlled test, run in a clean room. That makes it useful for spotting problems, but it is not a recording of what your real customers experience on their actual phones and connections. It's a model of speed, not a measurement of it.
Why it plateaus no matter what you do
There's a ceiling, and it's lower than most people think. A real store running real features carries weight that you cannot fully remove, because removing it would mean removing the functionality your business depends on. Your checkout, your essential apps, and parts of the Shopify platform itself all add load that no amount of optimization will zero out.
This is why I'm honest with people on the audit call: a perfect 100 is not a realistic target for a functional Shopify store, and anyone promising you 95+ is usually gaming the test rather than making your store genuinely faster. A score in the 70s on mobile is excellent for a working store. If you're sitting at 68 with real apps installed and you're trying to claw your way to 90, you may be fighting physics.
Does the score affect your Google ranking? Not how you think
This is the question that drives the most wasted effort, so here's the part worth reading twice. Google does factor page speed into ranking, through Core Web Vitals. But Google does not look at your Shopify admin score, and it does not look at the lab score in PageSpeed Insights either. It uses field data: the speed that real Chrome users actually experienced on your pages over the past 28 days.
Lab and field are two different things, and the gap between them is where all the confusion lives.
| Lab data | Field data |
|---|---|
| Your Shopify admin score and the top number in PageSpeed Insights | The "real users" panel in PageSpeed Insights and the Core Web Vitals report in Search Console |
| One simulated device, one simulated network, in a clean room | Thousands of your actual visitors on their own phones and connections |
| Great for finding and debugging issues | This is what Google uses to rank you |
So you can have an unimpressive lab score and still pass Core Web Vitals in the field, which is the outcome that actually helps your SEO. And the reverse happens too: a flattering lab score on a store whose real visitors are having a slow experience. Chasing the lab number can pull you away from the thing that matters.
Not sure if your store is actually slow?
I'll look at both your lab scores and your real-world field data and tell you which one to worry about. Free audit.
What to watch instead of the score
Open PageSpeed Insights and run a page that gets real traffic. If your store has enough visitors, a panel appears at the top with the heading about what your real users are experiencing. That's your field data. It shows your Largest Contentful Paint, your Cumulative Layout Shift, and your Interaction to Next Paint, each marked green, amber, or red.
Those three colours tell you more about your SEO and your customers than the lab score ever will. If they're green, your real visitors are having a good experience and Google sees it that way too, even if your lab number looks ordinary. If they're red, that's the real problem to fix, and it's the one worth your weekend.
The Core Web Vitals report inside Google Search Console shows the same field data across all your pages at once, grouped by issue. That's the dashboard I'd watch if I were you, not the admin score.
Why your score might be stuck even after real fixes
Pulling it together, there are really only a few reasons a score refuses to move after you've put in the work. The change was smaller than the natural bounce, so it got lost in the noise. You'd already reached the realistic ceiling for a store with your feature set. The fixes you made addressed lab flags that weren't your actual bottleneck. Or, the one I see most, there's leftover code from uninstalled apps still loading, so the store never really got lighter even though the app list got shorter.
In almost every one of those cases, the right move is to stop staring at the single number and look at the real-world picture instead. Measure your field Core Web Vitals, measure your actual load time on a real mid-range phone, and judge the work by those.
If you've been fighting a stuck score and you can't tell whether your store is genuinely slow or you're just chasing a misleading number, that's a good thing to hand to someone who looks at this every day. On a free audit I'll show you both sides, the lab and the field, and tell you honestly which one needs attention and which one you can stop worrying about.
Get a straight answer about your speed
A free 20-minute call. I review your store first and tell you what's real and what's just the number lying to you.