Most “best Shopify speed apps” lists are written by people who have never opened a theme’s JavaScript. So here is the honest version first: the stores we have measured at 1.2–1.4 seconds Largest Contentful Paint got there with zero speed apps installed, and the slowest stores we audit usually have two. An app can compress your images and prefetch your next page. It cannot remove the 900 KB of theme and app JavaScript that is the reason your store is slow.
This guide covers the three kinds of speed app that do something measurable, the ones to avoid, and the four causes of a slow Shopify store that no app touches. Before any of it, measure: our free Store Speed Test reads Chrome’s real field data for your domain — the same data Google ranks on — in the browser, with no sign-up.
Summary
- Image compression apps help, on stores that upload uncompressed images. Crush.pics and TinyIMG are the two we see work.
- Prefetch apps (Page Speed Booster) make the second page feel instant. They do nothing for the first.
- Script-deferral apps (Hyperspeed and similar) improve lab scores and can break the apps whose scripts they defer. Use with care, measure in the field.
- No app fixes theme JavaScript, app-injected scripts, an oversized hero or third-party pixels. That is theme work, and it is where the seconds are.
What “speed” means, and which number matters
Google measures three things in the field — from real Chrome users on your real store — and ranks and prices ads on them:
- Largest Contentful Paint (LCP): when the biggest thing above the fold has painted. Target under 2.5 s. This is the one most Shopify stores fail on mobile.
- Interaction to Next Paint (INP): how long the page takes to respond to a tap. Target under 200 ms. Theme and app JavaScript decide it.
- Cumulative Layout Shift (CLS): how much the page jumps while loading. Target under 0.1. Late-loading widgets and images without dimensions cause it.
Lab scores — the Lighthouse number, the “speed score” in the Shopify admin — are a simulation and they are what most speed apps optimise for. The field numbers are what customers experience and what Google uses. A speed app that raises your Lighthouse score from 41 to 68 and moves field LCP from 3.9 s to 3.7 s has done almost nothing.
What actually slows Shopify stores down
From the audits we run, in order of how often each is the main cause:
- Theme JavaScript. Premium themes load every section’s scripts on every page — the mega menu, the slider, the quick-view, the product recommendations — whether the page uses them or not.
- App scripts. Every app with a storefront presence injects a script through the theme app extension. Six to fifteen of them is typical; each one is a request, a parse and often a render.
- The hero image. A 3,000-pixel image served to a 390-pixel phone, often in the wrong format, often not the first thing the browser is told to fetch.
- Third-party pixels and tags. Meta, TikTok, Google, Klaviyo, Hotjar and a review widget, all loading before the first image.
- Fonts. Four weights of two families, loaded late, with the text invisible until they arrive.
Only the third and, partly, the fourth can be touched by an app. Our thirteen speed techniques go through the theme-level fixes for the rest.
The speed apps worth installing
Image compression: Crush.pics and TinyIMG
Crush.pics compresses the images already in your media library (lossy or lossless, your choice), keeps compressing new uploads, and can rename files and fill alt text. TinyIMG does the same with more SEO tooling bolted on. Either one produces a real, measurable gain on a store whose team uploads straight from the camera or the designer’s export — which is most stores.
What they cannot do: change which image the browser fetches first, serve a smaller image to a phone (that is the theme’s srcset), or convert the format the theme requests. Shopify’s CDN already serves WebP and AVIF when the theme asks correctly; if the theme does not, the compression app is polishing the wrong file.
Prefetching: Page Speed Booster
Page Speed Booster and apps like it prefetch the next page when a customer hovers or touches a link, so the collection-to-product click feels instant. It is free, light, and genuinely improves the browsing experience. It does not change LCP, INP or CLS on the landing page, which is the page Google measures and the page the ad click lands on, so it will not move your Core Web Vitals report.
Script deferral: Hyperspeed and Swift
Hyperspeed and Swift are the “optimise everything” category: they defer and reorder scripts, lazy-load images and sometimes rewrite how the theme’s assets load. On a bloated theme they can improve lab scores substantially and field LCP somewhat. The risks are real: a deferred script is one an app was relying on, and the symptom — a review widget that stops rendering, a cart drawer that opens empty — shows up a week later, on a phone, for a customer. If you install one, measure in the field for two weeks, test every app’s storefront feature, and do not install two of them; two deferral apps fighting each other is a finding we see regularly.
The speed apps to skip
- “Speed score” apps that display a number in the admin and change nothing.
- Lazy-load-everything apps on a theme that already lazy-loads; they often lazy-load the hero, which makes LCP worse.
- Any app that promises a 90+ score in one click. The score it means is the lab score.
Quick comparison
| App | What it does | Moves field LCP? | Risk | Install it when |
|---|---|---|---|---|
| Crush.pics | Compresses library images, alt text, renaming | Yes, on uncompressed uploads | Low | The team uploads camera or design exports |
| TinyIMG | Compression plus SEO tooling | Yes, same conditions | Low | You want image SEO in the same tool |
| Page Speed Booster | Prefetches the next page | No (second page only) | Very low | Always; it is free and light |
| Hyperspeed | Defers scripts, lazy-loads, reorders assets | Sometimes | Medium — can break app features | Bloated theme, no developer, measured in the field |
| Swift | Similar script and asset control | Sometimes | Medium | Same, never alongside another deferral app |
The four things no app can fix
1. Theme JavaScript loaded on every page
The fix is in the theme: sections load their own scripts only when the section is on the page, the slider library is replaced by CSS scroll-snap, the quick-view is loaded on first tap. On a custom theme this is how it is built from the start; on a premium theme it is a refactor. PawGuard’s theme carries no premium theme underneath and every non-essential block is stripped, which is how a store with 108 product images verifies at 1.4 seconds.
2. App scripts
Audit the theme app extensions and remove the apps you no longer use — uninstalling often leaves the script behind. For the apps you keep, load their scripts after the first paint where the app allows it. A store with fifteen storefront apps has a speed ceiling no theme work can lift.
3. The hero image
The first image should be the first thing fetched (a preload hint), served at the phone’s size (srcset and sizes set correctly), in a modern format, with dimensions declared so nothing shifts. Four theme changes, and typically the single largest LCP gain on a store.
4. Third-party pixels and fonts
Pixels move to a tag manager loaded after the first paint, or server-side. Fonts drop to two weights, preloaded, with the text visible while they arrive. Neither is an app’s job.
A sensible order of work
- Measure in the field with the Store Speed Test; note LCP, INP and CLS on mobile.
- Install Crush.pics or TinyIMG and Page Speed Booster. Low risk, real gains on the right stores.
- Uninstall unused apps and remove their leftover code.
- Fix the hero image, the fonts and the pixel loading in the theme.
- Refactor or replace the theme’s JavaScript loading — this is the project, and it is what our speed optimisation engagement is.
- Hold the result: every new app and campaign is measured before it stays. The performance retainer exists because speed decays.
If the number is bad enough that headless is on the table, read the twelve pros and cons first; most stores that think they need Hydrogen need a theme without the apps.
FAQ
Do Shopify speed apps actually work?
Image compression apps do, measurably, on stores with uncompressed uploads. Prefetch apps make the second page feel instant but do nothing for the first. Script-deferral apps can improve lab scores and break the apps whose scripts they defer. None of them removes the theme and app JavaScript that causes most slow Shopify stores.
What is a good page speed for a Shopify store?
Under 2.5 seconds Largest Contentful Paint in real field data (Chrome UX Report), under 200 ms Interaction to Next Paint, and under 0.1 Cumulative Layout Shift. The stores we build verify at 1.2–1.4 s LCP on production.
Why is my Shopify store slow even with a speed app?
Because the app cannot touch the causes: a theme that loads every section’s JavaScript on every page, six to fifteen app scripts injected through the theme app extension, a hero image served at 3,000 px to a 390 px phone, and pixels loading ahead of the first image. Those are theme and configuration work.
Should I uninstall speed apps before a speed optimisation project?
Usually, and we do it in the audit. Two script-deferral apps fighting each other is a common finding, and an image app that re-compresses already-optimised images adds nothing. Uninstall, then remove the leftover code the app left in the theme.
