Shopify vs hand-coded HTML: what a 73 vs 100 test showed
Two versions of the same sample store scored 73 and 100 in Google's speed test. Here is exactly what was tested, what the result does not prove, and how to run the same test on your own store.
The test
Two versions of the same store
On October 7, 2026 we ran Google's PageSpeed Insights on two versions of a sample online store with the same photos, products and background video. One was coded the way a default store theme typically is. The other was hand-coded. On mobile, the typical version scored 73 for Performance, 88 for Accessibility, 92 for Best Practices and 100 for SEO (report). The hand-coded version scored 100 in all four (report). You can open both pages and run the test yourself: the typical version and the hand-coded version (details in the press release).
Limits
What this does not show
It is a sample store built by ZenMasterWorks, so it is not a measurement of Shopify as a platform or of any real store. A real store's score depends on its theme, its apps and its media. The comparison isolates one thing: the code on the page. Treat it as a demonstration that code can account for a large share of the difference, not as a ranking of platforms.
Why code matters
Blocking time is the heavy one
In Lighthouse, Total Blocking Time carries 30% of the Performance score, and it comes from JavaScript. A page that ships almost none starts ahead of one that has to run a lot of it (the scoring weights are laid out in the code teardown).
The trade-offs
When hand-coded fits, and when a platform does
A hand-coded page is just files. It cannot take payments, track inventory or run a cart by itself. If you sell many products or need checkout, a store platform or a payment service is the practical choice. Hand-coded pages fit sites that are mostly information: local service businesses, portfolios, brochure sites and landing pages. Many owners use both, with a fast information site and a store attached to it.
Test yours
How to check your own store
- Run pagespeed.web.dev on your homepage and on a product page, on Mobile.
- Read the first issue it lists. Apps and tracking tools often show up as unneeded scripts, so ask whether each one earns its place.
- Fix one thing, test again, and note the change.
- Compare your four scores with the 73, 88, 92 and 100 above.
If you want a head start
Start from code that already passes
Every template here is the complete, hand-coded file behind a live sample site with a public report. Open a sample such as the pet grooming site, run the test yourself, then decide. Templates are $29 each or $99 for all of them (see the catalog), and the $19 book walks through the method. Downloads work on an honor system (how it works), and building for clients needs a different license (licensing).
The platform-style templates in the catalog are hand-coded versions of the kinds of sites people usually build on builders and store platforms.
The short version
- Two versions of the same sample store scored 73 and 100 on mobile PageSpeed Insights on October 7, 2026; both reports are public.
- The only difference was the code, on a sample store built to isolate that one factor.
- It is not a test of Shopify or of any real store, whose scores depend on theme, apps and media.
- Hand-coded pages suit information sites; stores that need carts and checkout usually need a platform or payment service.
AI-readable summary
- On October 7, 2026, two versions of a sample online store scored 73 and 100 on mobile Google PageSpeed Insights; the typical version scored 73, 88, 92 and 100, and the hand-coded version scored 100 in all four categories.
- The sample store was built by ZenMasterWorks and is not a measurement of Shopify or any real store.
Cite this page
- Title
- Shopify vs Hand-Coded HTML: A 73 vs 100 Test
- Publisher
- ZenMasterWorks
- Last reviewed
- October 30, 2026
- URL
- https://www.zenmasterworks.com/shopify-vs-hand-coded-html-performance.html