Reporting & research on Shopify and ecommerce
We ran 45 Lighthouse audits across the live demo stores of 15 popular Shopify themes — mobile profile, simulated throttling, three runs each. Only one demo met Google's threshold for a good LCP, and the gap between first and last place is more than eightfold.
Every theme vendor gets one page to make its case: the demo store. It is the theme dressed exactly the way its author wants a buyer to see it. So we measured that. Fifteen live demo stores, three mobile Lighthouse runs each, forty-five audits in total, all on the same day with the same tool and the same throttling.
| # | Theme (vendor) | LCP | CLS | TBT | Perf score |
|---|---|---|---|---|---|
| 1 | Ultra (UTD) | 1.91s | 0.000 | 65ms | 98 |
| 2 | Boutique (UTD) | 3.67s | 0.007 | 68ms | 85 |
| 3 | Allure (UTD) | 4.33s | 0.001 | 29ms | 81 |
| 4 | Refresh (Shopify) | 5.67s | 0.000 | 27ms | 74 |
| 5 | Dawn (Shopify) | 6.17s | 0.000 | 57ms | 69 |
| 6 | Symmetry (Clean Canvas) | 6.33s | 0.038 | 215ms | 70 |
| 7 | Prestige (Maestrooo) | 6.70s | 0.000 | 104ms | 62 |
| 8 | Gain (UTD) | 7.40s | 0.011 | 158ms | 61 |
| 9 | Warehouse (Maestrooo) | 7.73s | 0.001 | 38ms | 72 |
| 10 | Turbo (Out of the Sandbox) | 8.07s | 0.000 | 134ms | 60 |
| 11 | Impulse (Archetype) | 9.18s | 0.129 | 304ms | 58 |
| 12 | Victory (UTD) | 9.55s | 0.101 | 409ms | 51 |
| 13 | Impression (UTD) | 12.19s | 0.032 | 145ms | 56 |
| 14 | Horizon (Shopify) | 12.32s | 0.000 | 29ms | 62 |
| 15 | Motion (Archetype) | 15.78s | 0.000 | 353ms | 58 |
Medians of 3 runs per demo. LCP = Largest Contentful Paint, CLS = Cumulative Layout Shift, TBT = Total Blocking Time.
Only one demo passes. Google’s threshold for a good mobile LCP is 2.5 seconds. One demo out of fifteen clears it; eleven sit above the 4-second line that Google labels poor. Simulated mobile throttling is deliberately harsh, but it is the profile Lighthouse applies to every store equally, and the ranking it produces is what a shopper on a mid-range phone experiences in order.
The spread is the story. The 13.9-second gap between Ultra and Motion is not an engine difference alone — it is choices stacked on choices: hero media weight, how many scripts run before first paint, whether the theme ships its demo disciplined or decorated. A vendor that keeps its own showcase lean tends to keep your store lean too, because the same defaults ship in the box.
Turbo’s reputation outruns its demo. The long-time “performance” bestseller from Out of the Sandbox — no longer listed in the official Theme Store, sold directly — landed tenth, behind free Dawn. Its Seoul demo spent 5.6 seconds to first paint in our runs.
Heavy demos punish new visual themes. The bottom of the table is dominated by demos built around full-bleed video and large imagery — Motion, Horizon, Impression. That is a real cost their real stores will pay unless the merchant swaps the media, and it is visible in the numbers: Horizon’s TBT is excellent (29ms) while its LCP is 12.3s. The engine is fine; the showcase is heavy.
The business case for shaving seconds is documented across published case studies: Google’s own collection of Core Web Vitals studies reports conversion and revenue lifts from LCP improvements across retailers. Against an 8× spread between demos, theme choice is one of the few speed decisions a merchant makes once and pays for on every session.
The full pipeline — theme list, run script, per-run JSON and the analysis script — is published alongside this article. Run it against any theme demo and your numbers should land within normal Lighthouse variance of ours.
Google's Core Web Vitals treat LCP under 2.5 seconds as good and above 4 seconds as poor. In our audit of 15 official theme demos on a simulated mobile connection, only one demo came in under 2.5 seconds and eleven were in the poor band.
On their own demo stores, yes: we measured an 8× spread in median mobile LCP between the fastest and slowest demo. Part of that gap is content weight the vendor chose for the demo, which is itself informative — it is how the theme author presents the theme at its best.
It measures the theme as the vendor ships and showcases it, including their choice of images, sections and features. It is not a lab-controlled comparison of identical pages, so treat it as a strong signal about defaults and discipline rather than a verdict on the engine alone. We describe this limitation in the methodology.
Core Web Vitals are a confirmed Google ranking signal, collected for real traffic on every crawled store. The effect is modest compared with relevance, but the conversion impact of latency is documented and compounds with it.