Why Is My Shopify Store Slow to Load? What to Fix, in Order

A slow Shopify store is usually not one broken thing. It is a pile of small additions, each of which seemed harmless, and the fix is knowing which ones actually cost you time. That means starting with what gets measured.

The three numbers Google measures

Advice to get your store under three seconds is not wrong, but it is not what search engines assess. Google measures three Core Web Vitals, and each has a published threshold:

The detail that catches people out is the 75th percentile. You do not pass by being fast on average. You pass when three quarters of real visits meet the threshold, which means the slowest quarter of your traffic decides your result — older phones, worse connections, the customer on a train. Testing on your own laptop on office wifi tells you almost nothing about whether you pass.

What speed is worth, with an actual method attached

This page used to say a one-second delay can cut conversions by up to 7 percent. We cannot source that figure and have removed it. It circulates endlessly without a study attached, which is exactly the kind of claim that should make you suspicious of the page carrying it.

Here is one we can source. Google published a case study on Rakuten 24, a Japanese online store, which ran a month-long A/B test splitting traffic 50/50 between a Core Web Vitals-optimised landing page and the original. The optimised version produced a 53.37% increase in revenue per visitor, a 33.13% increase in conversion rate, and a 15.20% increase in average order value, with exit rate down 35.12%. The largest technical change was cutting CLS by 92.72% — stopping the layout jumping around, rather than raw load time.

Two honest caveats, because a single case study is not a law. This is one retailer, in one market, on one page type, and the improvement was on a page that had meaningful room to improve. Do not expect a 53% revenue lift because you compressed some images. What it does establish, with a real method behind it, is that layout stability is worth more than most owners assume — and CLS is the metric almost nobody optimises.

What Shopify already does for you

This section exists to save you money. Shopify's own documentation states that images are automatically optimized for your storefront with the Shopify Content Delivery Network, that it "automatically determines the best possible file format" and delivers modern formats such as WebP and AVIF where the browser supports them. You do not configure any of it and you do not pay extra for it.

If a proposal quotes you for setting up a CDN, enabling gzip, or turning on image compression on a Shopify store, that is a bill for work the platform did before you arrived. It is the single most common thing we see overcharged for in this category.

The fixes that actually remain, in order

1. Apps, and what they leave behind

Every app injects its own code into your pages, and it runs on pages that have nothing to do with the app. A reviews widget loads on your checkout. A currency converter loads on your blog.

The worse problem is uninstalling. Removing an app from your admin frequently leaves its snippets behind in the theme, so you keep the load cost after you stop getting the feature. When we audit a store, leftover code from apps the owner removed months ago is the most common single finding. It is also the cheapest thing to fix, because deleting dead code cannot break a feature nobody is using.

2. Images that are bigger than the space they sit in

Shopify compresses your images. It cannot decide that a 4000-pixel-wide product photo did not need to be 4000 pixels wide to fill a 600-pixel slot. Resize to the dimensions they actually display at before uploading, and the compression Shopify applies afterwards has far less work to do.

3. Theme weight and third-party scripts

Feature-heavy themes load code for sections you never used. Add a tag manager, a chat widget, two tracking pixels, and four custom font weights, and you have a slow page built entirely from things that individually seemed reasonable. Fonts are worth singling out: each weight and style is a separate file, and most stores load several they never display.

Measure first, and measure the same way twice

Run the store through Google PageSpeed Insights and check the Web Performance report in your Shopify admin. PageSpeed Insights shows both a lab test and, where Google has enough traffic data, field data from real visits. The field data is the one that matches how you are actually assessed, so if the two disagree, believe the field data.

Change one thing at a time and re-test. Fixing four things at once and seeing the score move tells you nothing about which one mattered, which means you cannot repeat it next quarter.

When a rebuild is cheaper than tuning

Sometimes the theme is the problem. A theme that has been modified by several developers over several years, with abandoned sections and layered overrides, can cost more to untangle than to replace. The signal we look for is whether fixes stay fixed: if changing one thing keeps breaking another, the code is working against you and patching it is throwing money at the symptom.

A clean rebuild on a modern theme keeps your products, customers, and orders — those live in Shopify, not in the theme — and it is usually faster to deliver than a long tuning engagement. That work is a large part of our Shopify development practice.

But we would rather tell you the truth about which one you need. If your store is slow because of images and three dead apps, that is a small job and we will say so. Send us the URL and we will run it and tell you what we find, free.