How to test shopify theme speedUse four tools together, not one score in isolation.
The best Shopify theme performance test checks synthetic lab data, real user vitals, request waterfalls, and the trendline Shopify sees inside the admin. Each tool answers a different question.
Core Web Vitals, field data, and mobile-first diagnostics
Google PageSpeed Insights
Run the homepage, one collection page, and your highest-traffic product page.
Check mobile first, then desktop. Most Shopify slowdowns show up faster on mobile.
Look at LCP, INP, CLS, render-blocking resources, and unused JavaScript before you look at a single headline score.
Waterfalls, request weight, and third-party script visibility
GTmetrix
Run the test from a consistent location so repeat checks are comparable.
Use the waterfall to find the requests that delay the first useful paint.
If one app script or image cluster dominates the waterfall, the issue is not abstract. It is visible.
Filmstrips, repeat views, and understanding what shoppers actually feel
WebPageTest
Use a mobile device profile and record a filmstrip if possible.
Watch how long it takes before the hero, product gallery, and add-to-cart area are usable.
Slow themes often reveal themselves as visual instability, late-loading images, or delayed interaction rather than just a weak synthetic score.
Theme-level monitoring inside the Shopify admin
Shopify Speed Score
Use it as a directional signal, not your only decision-maker.
Compare the score before and after large app installs, image uploads, or theme customizations.
If the score keeps slipping while merchandising complexity rises, your theme architecture may be running out of headroom.
Speed comparison tableShopFlow vs Dawn vs Impulse vs Turbo on the metrics speed-minded merchants actually watch.
These ranges are editorial profiles for configured storefronts, not guarantees. Real performance still depends on apps, media, and implementation quality. The goal is to show how much headroom each theme usually leaves before it starts to feel heavy.
ShopFlow
Best speed-first upgrade- Typical mobile score
- 82-92
- Typical LCP
- 1.8s-2.4s
- JS posture
- Minimal theme JS
- Risk
- Low theme bottleneck risk
- Best for
- Brands that want more conversion tooling without stepping into heavier premium-theme overhead.
Dawn
Best free baseline- Typical mobile score
- 78-88
- Typical LCP
- 2.0s-2.6s
- JS posture
- Very lean baseline
- Risk
- Low, but less built-in conversion UX
- Best for
- Merchants who want a free starting point and can live with a simpler merchandising layer.
Impulse
Promo-heavy premium theme- Typical mobile score
- 60-74
- Typical LCP
- 2.6s-3.4s
- JS posture
- Heavier promo modules
- Risk
- Medium once apps and rich media stack up
- Best for
- Merch-heavy brands that care more about promo flexibility than absolute speed.
Turbo
Agency power theme- Typical mobile score
- 64-78
- Typical LCP
- 2.4s-3.2s
- JS posture
- Feature-rich baseline
- Risk
- Medium in heavily customized builds
- Best for
- Experienced teams that want deeper controls and can manage the added complexity carefully.
| Theme | Typical mobile score | Typical LCP | JS posture | Theme bottleneck risk | Best for |
|---|
ShopFlowBest speed-first upgrade | 82-92 | 1.8s-2.4s | Minimal theme JS | Low theme bottleneck risk | Brands that want more conversion tooling without stepping into heavier premium-theme overhead. |
|---|
| 78-88 | 2.0s-2.6s | Very lean baseline | Low, but less built-in conversion UX | Merchants who want a free starting point and can live with a simpler merchandising layer. |
|---|
ImpulsePromo-heavy premium theme | 60-74 | 2.6s-3.4s | Heavier promo modules | Medium once apps and rich media stack up | Merch-heavy brands that care more about promo flexibility than absolute speed. |
|---|
| 64-78 | 2.4s-3.2s | Feature-rich baseline | Medium in heavily customized builds | Experienced teams that want deeper controls and can manage the added complexity carefully. |
|---|
Editorial note: these ranges assume reasonably configured stores. Third-party apps, image choices, custom code, and traffic region can materially change the final result.