To improve Shopify speed, find the failing metric first, then fix the layer behind it. Open Online Store, then Themes, and read LCP, INP and CLS on mobile, using real visitor data. On mobile, 76% of Shopify sites pass Core Web Vitals, so the platform is seldom what holds a store back. Remove unused apps and tags first, which Shopify calls the highest-impact optimisation. Then fix the main image, the remaining JavaScript and slow Liquid, in that order. Be wary of speed apps and cheap services: Shopify documents ones that change the test score and leave real visitors no better off.
Most slow Shopify stores are slowed by what was added to them
A slow Shopify store is usually slow because of its apps, tags, images and theme code, and seldom because of the platform. The 2025 Web Almanac found that 76% of Shopify sites pass all three Core Web Vitals on mobile, based on real visitor data from Chrome.
| Mobile, 2025 | Sites measured | Good LCP | Good INP | Good CLS | Pass all three |
|---|---|---|---|---|---|
| Shopify | 567,932 | 86% | 90% | 92% | 76% |
| WooCommerce | 995,782 | 39% | 88% | 85% | 35% |
These figures mix small and large stores, and a third-party tool detected the platform, so read them as a broad picture. They still answer the first question. If your store fails, about three in four Shopify sites are passing on the same platform, so look at your own store before blaming Shopify.
Shopify already handles much of what generic speed checklists recommend. Its platform documentation and Help Center list a CDN, compression, minified CSS and JavaScript, image optimisation and browser caching set to one year. You do not need to buy or switch on any of those. What you control is everything sitting on top of them.
Measure real visitors first, because the old speed score is gone
Start with Shopify's Web Performance dashboard, which shows how real visitors experienced your store over the past 30 days. It replaced the Lighthouse-based Speed Score in 2024, so advice about raising your Shopify speed score is out of date.
- Open the summary. Go to Online Store, then Themes. The performance summary rates LCP, INP and CLS as Good, Moderate or Poor and shows sessions by device type.
- Open the full reports. Go to Analytics, then Reports. Search "contentful" for LCP, "next" for INP and "cumulative" for CLS. Staff need the Reports permission.
- Filter to your main device. Use the sessions split to pick the device most of your visitors use. Then read the Page Type report to see which template fails and the Page URL report to see which pages.
- Read the Over Time chart. It marks events such as app installations, theme updates and new code, so a sudden drop can often be matched to a named change.
| Metric | What it measures | Good | Poor |
|---|---|---|---|
| LCP (Largest Contentful Paint) | When the main content has loaded | 2.5 seconds or less | Over 4 seconds |
| INP (Interaction to Next Paint) | How fast the page responds to a tap or click | 200 ms or less | Over 500 ms |
| CLS (Cumulative Layout Shift) | How much the layout jumps while loading | 0.1 or less | 0.25 or more |
These figures are the 75th percentile, so three in four visits were at the number shown or better. Data can lag by up to 36 hours and is kept for 90 days. A new store or one in private mode may show no data. Shopify's performance team suggests a store with fewer than 100 users a day should work on traffic first, and that a store with fewer than 20 views a day should group results by week or month, because daily numbers swing.
Use Google's tools as a cross-check
Search Console's Core Web Vitals report groups similar URLs, separately for mobile and desktop, and gives each group the status of its slowest metric. PageSpeed Insights shows real visitor data from the previous 28 days, plus a lab test.
The lab test is one simulated load on a mid-tier phone, a Moto G4, run from North America, Europe or Asia. It is useful for debugging and it does not represent your buyers. If you sell mostly in India, your field data already reflects your customers' phones and networks, so trust it over the lab number and test by hand on a mid-range Android phone using mobile data.
A low Lighthouse score on its own is no emergency. Shopify's minimum for Theme Store themes is an average score of 60 across the home, product and collection pages.
The failing metric tells you which layer to fix
Each metric points to a different layer, so read the report before anyone touches code. Google's guidance counts a page as passing only when all three metrics meet their targets at the 75th percentile.
| What the data shows | Usual layer | First test |
|---|---|---|
| LCP is slow | Main image, blocking scripts, pop-ups | Check the Largest Contentful Paint element in PageSpeed Insights |
| Server response (TTFB) is over 0.8 seconds | Liquid in the theme | Run Shopify Theme Inspector |
| INP is slow | JavaScript from the theme, apps or tags | Block one script's domain in DevTools and compare |
| CLS is high | Images without set sizes, banners that appear late, font swaps | Watch the page load on a phone |
| One template fails | Sections and apps on that template | Read the Page Type report |
| A sudden drop on one date | A recent change | Read the event markers on the Over Time chart |
Shopify also says a customer's device, network and location are out of your hands. If the overall metric is good, a few pages or dates in the Moderate or Poor range are less of a concern.
Remove code before you optimise anything
Removing an app or tag is the cheapest speed fix, because every script you remove stops competing for the browser's main thread. Shopify's script audit guide calls this the highest-impact optimisation for responsiveness and overall performance, and calls the usual pattern "death by a thousand cuts": tags and apps added over time that keep running after they stop being needed.
- List what is running. Run Lighthouse and open its treemap to see JavaScript grouped by origin and size. Settings, then Apps now shows each app's latest activity, which helps you find apps nobody uses.
- Check three templates. In Chrome DevTools, open More tools, then Coverage, and reload the home, collection and product pages. A script idle on one page may matter on another.
- Measure one script's cost. Block its domain with Network request blocking, reload and compare transfer size and request count. Then record a Performance trace and compare scripting time in the Bottom-Up tab, grouped by URL.
- Ask who still needs it. Shopify's list of usual bloat covers overlapping analytics, A/B tools that inject blocking JavaScript, chat widgets, review apps loading on every page, email capture dialogs, social embeds and retargeting pixels.
- Delete the leftovers. Shopify warns that uninstalling an app does not automatically remove its code. Search
layout/theme.liquidandsnippets/for the app's name and script domains, along with anyrendertags that call those snippets. Work on a duplicate theme.
Indian stores should put COD forms, one-click checkout pop-ups, pin code checkers and WhatsApp widgets through the same test. Some of them will earn their cost. Measure each one, then decide whether it pays for itself. Our app audit guide covers the commercial side of that decision, and our guide to developer access covers giving someone safe access to do this work.
Scripts you keep should load only where they are needed. Shopify's examples load a review widget only on product pages, using {% if template.name == 'product' %}, and a chat widget after the first scroll. It notes that chat widgets often load hundreds of kilobytes of JavaScript on every page.
Clean out the tag manager
Audit the container and delete tags that are unused or that duplicate Shopify's built-in analytics. Shopify says tag managers that inject more resources create dependency chains that delay rendering. Remove any A/B test anti-flicker snippet when no test is running: one Shopify example shows the first paint near 2.75 seconds when the HTML had finished before 1.2 seconds.
Moving tags into a custom pixel is a tracking decision, and you should not count on it as a speed fix. Shopify's custom pixels page describes the sandbox in terms of security and data control, makes no performance claim, says not all pixel functionality works there, and says custom pixels are unsupported by Shopify. Our GA4 setup guide covers the tracking side.
Fix the main image next, because it usually sets LCP
If LCP is slow, check how the main image loads before anything else. On a product page or home page it is usually the largest element in view, and small changes to how it loads can move the metric. Shopify's team found that Shopify sites that lazy load this image have a median LCP 1.0 second slower, in an article dated January 2024.
- Never lazy load it. Shopify's image guidance calls this one of the most common and damaging patterns in themes. The
image_tagfilter loads images eagerly in sections 1 to 3 and lazily from section 4 on, so a hero moved down the page can change behaviour without anyone noticing. - Mark it as the priority. Add
fetchpriority="high"to that one image. Shopify's team tries to keep preloads and high-priority items to 2 or fewer each, because overusing them makes pages slower. - Use a real image tag. A hero set as a CSS background is found late by the browser. Shopify recommends
image_tagin Liquid so the image is discoverable in the HTML. - Remove fades on the hero. In one Shopify example, the image finished downloading around 7.6 seconds and LCP landed at 8.7 seconds because of a transition.
- Size images to how they display. Set width and height, or an aspect ratio, so the browser reserves the space. That also helps CLS.
- Check pop-ups and cookie banners. A pop-up that appears before any interaction can become the LCP element, and a cookie banner larger than your main content can do the same. Most pop-up apps have a setting to delay until the visitor interacts.
A hero video needs the same discipline. On Vaahe, we kept the hero video and imagery from costing the site its load time, because a slow store undercuts a premium claim.
Slow interactions come from too much JavaScript
Poor INP means the browser is too busy running scripts to respond when someone taps. Shopify's Help Center puts it plainly: most likely you have too much JavaScript from theme code, app code or tag managers, and you should clean up the worst offenders.
After the removals above, Shopify's theme best practices give the next moves. Use defer on non-critical scripts, and load a module with dynamic import() only when someone uses that component. Our INP guide covers how to profile a slow tap on a real phone and how to rework a cart drawer.
Check what you defer. Shopify's distinction is useful: delaying a chat widget is an improvement, and delaying the theme's own variant selector makes the page worse.
A slow server response points to Liquid and page size
If the wait before anything arrives is long, the cause is usually in how the page is built, and front-end fixes will not shorten it. Google rates a time to first byte of 800 ms or less as good. Shopify's guidance says Liquid rendering time directly determines it. PageSpeed Insights reports this figure, but on a streamed page it covers only the head, and the rest of the Liquid time appears as document download, so check both.
- Profile the Liquid. Shopify Theme Inspector for Chrome shows render time as a flame graph on your own store, or on stores where you have collaborator access. Shopify rates deeply nested loops as high impact, because they grow quadratically with your catalogue.
- Use JSON templates. Shopify's streamed HTML applies only to pages rendered from JSON templates. It also stops working if
content_for_headeris filtered or wrapped, so keep it as a plain tag. - Shrink the HTML. One Shopify example had 6,800 DOM nodes, of which 2,868 belonged to the mobile menu and 2,223 to the desktop menu. Large mega menus are worth checking.
- Cut sections and products per page. Shopify advises against adding too many sections to a template and suggests pagination for large collections.
If server response worsened at the same time as other Shopify sites, Shopify says to suspect a temporary issue on its side. It calls these rare and says they usually recover quickly.
Test the theme before replacing it
To see how much your theme and its additions cost, run Lighthouse in DevTools on a preview of an unedited free theme and compare it with your live theme. Shopify notes that previews are not streamed, so read the gap as a rough guide. Shopify suggests a performance-optimised theme if the current one causes significant slowdowns, and names Horizon and Online Store 2.0 themes. Our guides on moving to Horizon and the signs a store needs a rebuild cover that decision. A new theme will not help for long if the same apps go back on top of it.
Speed apps and cheap speed services can fake the result
Treat any promise of a higher score with suspicion, because Shopify documents services that change the test result and leave real visitors with the same store. Its page on fake performance apps says Shopify has removed most of these apps from the App Store, and that the same techniques are now sold on Fiverr and Upwork.
It describes three methods:
- LCP hijacking. A transparent image or blank element covers the screen and paints early, so the browser records it as the main content. The real hero loads as late as before.
- Test tool detection. A script spots Lighthouse, GTmetrix or an emulated Moto G and serves those tools a stripped page.
- Delayed work. Scripts are held back until the first scroll or tap. The score improves, and the page can feel broken when a customer first uses it.
You or your developer can check for these with Shopify's own tests:
- Run PageSpeed Insights, then Lighthouse in Chrome DevTools in Incognito. A difference of 20 points or more is a warning sign.
- Look at the PageSpeed Insights screenshots. A broken-looking page is another.
- Open the Largest Contentful Paint element diagnostic. It should name your hero and never a transparent image or empty div.
- Leave a page idle, then move the mouse once and watch for a burst of script requests. Then try the cart drawer and variant selector straight away.
- Compare with the Web Performance dashboard. Shopify says real users are usually faster than lab tests, so a great lab score beside poor field data needs explaining.
To remove the code, check layout/theme.liquid, snippets/meta-tags.liquid and snippets/social-meta-tags.liquid for references to Lighthouse, GTmetrix, Moto G or Linux x86_64, and for document.write(). Test on a duplicate theme and confirm what the code does first, because Shopify notes some uses are legitimate.
Judge any app or service on its results in your dashboard. A paid app that offers to minify CSS and JavaScript adds little, since Shopify does both already. The Built for Shopify rules say an app must not reduce your storefront Lighthouse score by more than ten points, which is a limit per app and leaves the total for your whole stack open. Before paying for speed work, ask for three things in writing: what will be removed or changed, which theme it will be tested on first, and which dashboard metric should move.
Stop speed work once your main templates pass on mobile
Fix speed until your highest-traffic templates are Good on your main device, then move the budget to conversion work. Past that point, the next gains are more likely to come from what the page says and asks of the visitor, and our CRO audit framework is the place to start.
The case for speed is real and easy to overstate. A study by Deloitte and 55, commissioned by Google, covered 37 brands and over 30 million sessions at the end of 2019. For retail sites, a 0.1 second mobile improvement across four speed metrics went with 8.4% more conversions and 9.2% more spend. Those were leading brands, one of the metrics has since been deprecated, and the improvement had to happen on every page, so use it as a direction and keep it out of your forecasts.
On search, Google says Core Web Vitals are used by its ranking systems. It also says good scores do not guarantee top rankings, and that it always seeks to show the most relevant content even when page experience is weak. It still highly recommends reaching good scores.
What to do before the sale, and how to confirm it worked
Black Friday is on 27 November 2026. Our BFCM readiness plan treats mid November as the usual code freeze, so do the removals and image fixes now and leave any theme change until after the peak.
Do not judge a fix from the summary score. It covers 30 days, so it moves slowly, and Shopify warns that metrics may not change right after an edit. Use the Over Time report with daily or weekly grouping, depending on traffic. In Search Console, Start Tracking opens a 28-day monitoring session, and an issue counts as fixed only if it appears on no URL during that window.
Speed slips again each time someone adds an app or a tag, so checking it belongs in routine upkeep. Our maintenance retainer guide sets out what that should include, and it is part of what we do through our Shopify maintenance service. If you want a second opinion on your own numbers, you can reach us through our contact page.
Questions
Why is my Shopify store so slow?
Usually because of what has been added to it: apps, tracking tags, large or lazy-loaded images, and theme code. The 2025 Web Almanac found 76% of Shopify sites pass Core Web Vitals on mobile, so the platform is seldom what holds a store back. Open the Web Performance dashboard to see which metric fails, then check the layer behind it.
How do I check my Shopify store speed?
Go to Online Store, then Themes, for a summary of LCP, INP and CLS from real visitors over the past 30 days. For detail, go to Analytics, then Reports, and search "contentful", "next" or "cumulative". Search Console and PageSpeed Insights give a useful second view. Read their real visitor data ahead of the lab score.
What is a good Shopify speed score?
Shopify no longer has a single speed score: it replaced it with three measures in 2024. Good means LCP of 2.5 seconds or less, INP of 200 ms or less and CLS of 0.1 or less, at the 75th percentile of real visitors. A Lighthouse score is only a lab reading: Shopify's minimum for Theme Store themes is an average of 60.
What is the best Shopify speed optimization app?
No single app can fix a slow store, because the usual cause is code that needs removing. Be careful with this category. Shopify documents apps and services that fake the test result, and it has removed most of those apps from the App Store. Judge any tool by the change in your Web Performance dashboard, and test it on a duplicate theme first.
Are Shopify speed optimization services on Fiverr safe?
Some are not. Shopify says techniques that cheat performance tools are sold on Fiverr and Upwork. They can raise a lab score while real visitors see no gain, and some leave the page unresponsive. Ask what will be removed or changed, insist on a duplicate theme, and judge the result on real visitor data.
Will a new theme make my Shopify store faster?
It can, if your current theme causes significant slowdowns. Shopify names Horizon and Online Store 2.0 themes as optimised for performance. Compare a preview of an unedited free theme with your live theme first, as a rough guide. If most of the weight comes from apps and tags, a new theme will not help for long once they are reinstalled.
How long do Shopify speed fixes take to show up?
Allow up to 36 hours for new data and up to 30 days for the summary to reflect a fix in full, because it covers the past 30 days. Use the Over Time report to see a change sooner. PageSpeed Insights field data covers the previous 28 days, and Search Console runs a 28-day monitoring session once you start tracking a fix.
Does Shopify store speed affect Google rankings?
Yes, though relevance comes first. Google says Core Web Vitals are used by its ranking systems and highly recommends good scores, but it also says good scores do not guarantee top rankings and that it always seeks to show the most relevant content. The stronger reason to fix speed is the visitor who reaches your store on a phone and has to wait for it.
