Shopify site speed and the apps you install

Short answer: will a video app slow my store down?

A badly built one will. A well-built one should be a rounding error — and you should not take either on trust, because it takes about two minutes to measure.

The honest version is that "does it slow my site down" is the wrong question, because every app costs something. The right question is how much, and is it visible in the metrics Google and Shopify actually grade you on. Run PageSpeed Insights before installing, install and configure the app as you really would, run it again, and compare Largest Contentful Paint and total blocking time. That delta is your answer, on your theme, with your traffic — not the vendor's.

For Quinn specifically, the design targets are published rather than asserted: widget bundles under 10 KB per component, loaded only when your chosen design uses them, and video encoded to a VMAF 93 perceptual target with per-scene resolution and bitrate, written faststart and served from Cloudflare with below-fold clips lazy-loaded. Every one of those is checkable in DevTools, and the free plan means checking costs nothing.

Where Shopify stores actually lose speed

1. App JavaScript. Each app adds scripts that must be downloaded, parsed and executed — and parse/execute cost is worse than download cost on mid-range phones, which is what most of your traffic uses.
2. Uncompressed media. Images and video straight from a camera or phone are enormous relative to what a page needs.
3. Render-blocking resources. Anything in the critical path delays first paint.
4. Theme-code injection. Apps that write into your theme leave code behind and break on theme updates.
5. Third-party embeds. A YouTube player or chat widget pulls in an entire external stack.

Note the order: JavaScript weight usually costs more than the video itself, once video is properly compressed. Most app-shopping advice ignores the bundle entirely.

The bundle problem nobody checks

Most widget-based apps ship a single monolithic bundle containing every feature they offer to every merchant. If the app supports carousels, stories, grids, floating players, pop-ups and six themes, you download all of it — even if your store uses one carousel. Your customers pay, in bytes and parse time, for features you have never enabled.

It is an understandable way to build: one bundle is simpler to ship and cache. But it means the app's page-speed cost is set by the size of its feature list rather than by what you actually use, and it grows every time the vendor adds something for someone else.

How Quinn's widgets are built

We took the opposite approach, and it is why the numbers hold up:

• Built in Svelte. Svelte compiles components to plain JavaScript at build time rather than shipping a framework runtime to the browser. There is no React or Vue runtime tax on your storefront — the shipped code is close to the code that actually does the work.
• Every component decoupled. Each widget component is independently loadable rather than baked into one artifact.
• Loaded only when needed, in parallel. Your store downloads only the components your chosen design actually uses, fetched concurrently rather than in a chain.
• Bundles under 10 KB. Each one. That is small enough that the widget is not a meaningful line item in your performance budget.

The consequence is the part we care about most: you get a full design system with effectively no performance penalty for using it. You can restyle the widget to match your store — layout, type, colour, behaviour — and because components load on demand, a more elaborate design does not translate into a heavier page. Design complexity and load cost are decoupled, which is normally a trade-off merchants are forced to make.

Auditing any app before you install it

This works for any Shopify app, not just video ones:

1. Run PageSpeed Insights first. Record LCP, TBT and total page weight as your baseline.
2. Install the app and configure it as you actually would. A default install understates real cost.
3. Re-run and diff. Anything more than a marginal LCP change deserves a conversation with the vendor.
4. Open DevTools → Network → JS. Sort by transferred size and find the app's scripts. This is the number vendors rarely publish.
5. Check whether it injects theme code. An app using the modern app-embed model does not edit your theme — and no theme-code injection is one of the requirements behind the Built for Shopify badge.
6. Check Online Store → Themes → Store speed after a few days of real traffic.

What good looks like

The other half of the story

A light widget serving heavy video is still a slow page. Quinn encodes every video to a VMAF 93 perceptual target with resolution and bitrate varying per scene, writes it faststart so playback begins on the first bytes, and serves it from Cloudflare — deliberately as a single stream rather than adaptive bitrate, because manifest round-trips and quality oscillation both hurt on short reels. The full encoding write-up →

Questions