Shopify speed work,
measured on your store.
A Lighthouse screenshot is a lab result on a machine that is not your customer's. We instrument a real Chrome against your real storefront, get LCP, CLS and INP the way Google records them, and fix what the numbers actually point at — which is regularly not what the audit tools flag first.
Why most Shopify speed reports are useless
Three things that go wrong
A score is not a metric. "Performance 47" tells you nothing about which of your customers waited. Google ranks on field data at the 75th percentile — the three Core Web Vitals — and a lab score can move without any of those moving.
The audit ran on the wrong page. Homepage audits are what get sold. Your revenue is on product and collection pages, which are heavier, more app-laden and measured separately.
The recommendations are generic. "Eliminate render-blocking resources" is not a finding, it is a category. The finding is which resource, loaded by which app or theme file, costing how many milliseconds.
What we actually do
- Measure properlyReal Chrome driven over the debugging protocol against your live storefront — home, collection, product and cart — on throttled mobile as well as desktop. We keep the raw traces, so every claim we make afterwards can be pointed at.
- Attribute the costEvery script gets attributed to the app or theme file that loaded it, with a millisecond cost attached. This is the part that changes decisions: it is much easier to drop a $19/month app when you can see it costs 900ms on every product page.
- Fix the themeImage sizing and formats, font loading, above-the-fold CSS, deferred and conditional scripts, and the layout shifts caused by elements that arrive late without reserved space.
- Cut the app taxThe largest single win on most stores. Apps that are unused, duplicated, or doing something a section of Liquid does for free — and a plan for removing them without breaking merchandising.
- Re-measure and hand overThe same instrumentation, the same pages, before and after, side by side. Plus what to watch so it does not drift back the next time someone installs something.
The honest part
Some stores are slow because of the theme, and those we can transform. Some are slow because of fifteen apps the business genuinely depends on, and the honest ceiling there is lower — we will say so after the measurement rather than after the invoice. And speed is one ranking input among many: it will not, on its own, move you up for a term you have no authority on.
The three metrics, in plain terms
| Metric | Good (p75) | What a shopper experiences |
|---|---|---|
| LCP | ≤ 2.5s | How long until the main product image and title are actually on screen. |
| CLS | ≤ 0.1 | Whether the page jumps while they read — the reason people tap the wrong variant. |
| INP | ≤ 200ms | How long the page feels frozen after they tap add-to-cart. |
INP is the one that catches Shopify stores now that it has replaced FID: it measures every interaction, not just the first, so a heavy variant picker or a slow cart drawer shows up where it previously did not.
Questions we get asked before signing
Can you guarantee a 90+ score?
No, and be careful with anyone who does. A lab score can be gamed in ways that do nothing for a real shopper. What we commit to is a measured before-and-after on your real store pages, with every change attributed to a number.
Will you break my apps?
Nothing is removed without your sign-off. We give you the cost of each app in milliseconds and you decide what stays. Removals happen one at a time, with the storefront checked between each.
How much can I realistically gain?
On a marketplace theme with an accumulated app stack, cutting largest-contentful-paint by a third to a half is common, because most of it is script the page never needed. On an already-lean custom theme the remaining wins are smaller and we will tell you that before you commit.
Do you also do this for headless stores?
Yes — and the diagnosis is usually different. Headless builds tend to be fast on first paint and slow on interaction, so INP is where the work lands.
Is speed really a ranking factor?
Core Web Vitals are a confirmed but modest signal — a tiebreaker between comparable pages rather than a way to outrank a stronger domain. The stronger argument for fixing it is conversion, not ranking.
Send us your store URL
We will measure it properly before the call and come to it with numbers — the real LCP, CLS and INP on your product pages, and what is causing them.
Book a call →More on Shopify
Custom theme development
When the theme itself is the ceiling, optimisation has a limit. This does not.
Shopify app development
Apps built to the performance budget instead of retrofitted into it.
The real Built for Shopify criteria
The performance thresholds Shopify actually publishes, with the myths removed.