Shopify site speed and SEO: what actually matters

The short answer

Speed work on Shopify pays in a specific order: audit the app list first, because injected scripts are where most stores lose their performance, then fix the Largest Contentful Paint element on the templates that earn money, then leave the micro-optimisations alone. Speed is a real ranking signal with a modest weight, so the work belongs after relevance and authority in the queue, and the measurements that matter are field data rather than lab scores.

The real weight of speed as a ranking factor

Core Web Vitals are used by Google's ranking systems, and have been since the page experience update rolled out in 2021. The weight of the signal is real but frequently overstated. Google's own documentation says it plainly: search always seeks to show the most relevant content, even when the page experience is sub-par. A site with a poor speed score but strong relevance and authority will outrank a fast site with weak content on most commercial queries.

The hierarchy is relevance first, authority second, then technical health, which is where speed sits. Speed work matters, but it matters after the foundational work is done. A Shopify brand that spends three months optimising Largest Contentful Paint while its collection pages have no content and its schema is broken has put the effort in the wrong place.

The commercial case for speed is stronger than the ranking case, and it deserves saying that way round. A slow store loses buyers directly, on mobile above all, and for a premium brand the sluggish page reads as a quality signal in itself. Fix speed for the customer, and take the ranking margin as the side benefit.

The three numbers, and where they need to be

The thresholds are Google's own, assessed at the 75th percentile of real visits, which carries a practical consequence: the score that counts comes from field data, the Chrome UX report numbers in PageSpeed Insights and Search Console, not from a lab run on office wifi. A store can pass in the lab and fail in the field on mobile networks, and it is the field that ranks.

Where Shopify stores slow down

Shopify's hosted infrastructure is fast by default. The CDN is global, server response times are strong, and the base platform imposes less overhead than a typical self-hosted install, which means server-side performance is one problem Shopify merchants can cross off entirely.

The speed problems on Shopify are almost always app-related. Each app adds JavaScript to the page, and many load their scripts on every page whether they are relevant there or not. A store with fifteen active apps is often carrying five or six scripts from apps that are installed but doing no commercial work, and uninstalled apps sometimes leave script tags behind in the theme.

The second tier of culprits is the tracking stack: tag managers carrying years of accumulated pixels, session recorders, heatmaps and attribution tools, each individually justified and collectively heavy. Every script in the stack should have an owner who can say what decision it feeds, and the ones without an answer are weight.

The app audit, step by step

1. List every installed app, and against each record what it does, who uses its output, and the last time anyone did.
2. Uninstall the ones with no owner and no output. On most established stores this is a quarter of the list.
3. For the keepers, check where their scripts load. An upsell app needed on product pages has no business loading on the blog, and many apps expose page-scoping settings nobody has opened.
4. After uninstalling, check the theme for leftovers: orphaned script tags and snippets from departed apps linger in theme.liquid and keep loading for years.
5. Re-measure the money templates in the field data a few weeks later, because that is the number that was ever worth moving.

Images, and the LCP element

Shopify's CDN serves uploaded images in modern formats like WebP and AVIF wherever the visitor's browser supports them, at sizes appropriate to the device requesting the page. For most stores, the image optimisation Shopify does automatically is already better than a manual setup would manage.

The image problems that persist sit in lazy loading, above-the-fold image sizing and identifying which element is the Largest Contentful Paint. If the LCP is an image, usually the hero on the homepage and the featured image on product pages, that image should be excluded from lazy loading and prioritised for early load, which is theme-level work a competent developer can do in an afternoon.

The one image mistake the CDN cannot fix is dimension: a 4,000-pixel original uploaded for a 600-pixel slot still costs processing even after Shopify resizes it, and enormous uploads accumulate in stores where several teams add imagery. A sensible upload standard, set once, costs nothing.

What not to spend on

Score-chasing is the trap. A Lighthouse score of 100 is not a commercial outcome, and the last twenty points typically cost more than the first sixty delivered, pursued through minification rounds, font subsetting and critical-CSS gymnastics that shave milliseconds no customer feels. When the field data is green across the money templates, the speed work is done and the budget belongs back in content and authority.

Be equally wary of speed apps that promise automated optimisation, since an app added to reduce the weight of apps starts from a difficult position, and several achieve their scores by deferring scripts in ways that break the tools other teams rely on. The durable wins on Shopify are subtraction: fewer scripts, scoped loading, a disciplined theme. Everything else is maintenance.

Questions we hear about Shopify speed