Reviewed November 5, 2026

How to Improve Your Shopify Speed Score Without a Rebuild

You can improve a Shopify store's speed materially without a rebuild by fixing, in order, the largest contentful paint image, the JavaScript your apps inject, images lazy loaded that should load first, and leftover code from uninstalled apps. Shopify's web performance dashboard shows LCP, INP, and CLS from real visits, so measure there before and after each change rather than chasing a single lab score.

A store owner opens the speed report, sees a number they dislike, and gets quoted for a theme rebuild by the first developer they ask. Sometimes a rebuild is right. More often the store is carrying four specific, fixable weights, and removing them buys most of the gain at a fraction of the cost and none of the launch risk.

Work the list in order. Each fix is measurable on its own, which is the other reason to skip the rebuild: when everything changes at once, nobody learns which change mattered.

The hero image is usually the LCP

Largest contentful paint on a Shopify homepage or product page is usually the hero or main product image. The common faults are a desktop sized file served to phones, and lazy loading applied to the one image that should load first. Serve the image at the size it displays, use modern formats the platform offers, and make the LCP image load eagerly with high fetch priority while everything below the fold lazy loads.

Check the timing breakdown in a lab test to confirm the image is the LCP element before optimizing it. Sometimes the LCP is a text block waiting on a webfont, and stores have resized images for a week chasing a font problem. Name the element first.

Apps write checks your theme cashes

Every app that touches the storefront can add JavaScript, and app scripts are a leading cause of poor interaction response, measured as INP. Audit in the theme code and the performance report which apps inject on which templates, then remove or replace the ones whose function does not earn their weight. Reviews, chat, and personalization apps are frequent heavyweights on product pages.

Deleted apps leave residue: script tags, snippet includes, and app blocks' orphaned code in theme files. Search the theme for the names of apps you uninstalled months ago. Finding their code still loading is common, it is pure cost with no function attached, and removing it is the closest thing speed work has to free money.

Fonts, carousels, and the third tier

Webfonts block text rendering when loaded carelessly. Limit the weights you actually use, and let the browser swap text in rather than holding the page for the perfect face. Hero carousels with three slides ask the browser to prioritize three LCP candidates at once; a single static hero almost always paints faster and converts no worse in most honest tests.

None of these fixes needs a new theme. They need an afternoon, a duplicate theme to work in, and the discipline to ship one change at a time and read the real user data a week later. Keep the rebuild quote for the day the theme itself, not its cargo, is the problem.

Frequently asked questions

What is a good LCP for a Shopify store?

Google's good threshold is 2.5 seconds or less, met by at least 75 percent of visits, measured separately on mobile and desktop. Shopify's web performance dashboard reports your real visitor LCP against that line by page type.

Why is my speed score fine on desktop and bad on mobile?

Mobile visits run on slower processors and networks, and Google weighs mobile heavily. Heavy JavaScript and oversized images that a desktop shrugs off dominate mobile timings. Always read the mobile figures first.

Do I need to remove apps to get faster?

Not all of them, but each storefront app should justify its script weight with revenue or a function nothing else covers. The audit usually finds two or three apps, plus leftover code from already deleted ones, that can go without any shopper noticing.

How long before speed changes show in the data?

Lab tests respond immediately. Field data, the real visitor figures in Shopify and Search Console reports, is a rolling 28 day window, so genuine improvement phases in over two to four weeks. Judge a change by lab first, field second.