How fast are US online stores on a phone?
We ran Google’s phone test on the home pages of 367 US brands that sell direct to shoppers, and read Google’s real-visitor data where it exists. These are the numbers.
7IT SolutionsPublished 367 stores
What we found.
- 34median score out of 100 on Google’s phone test
- 11.7 smedian time to the main content on a mid-range phone
- 0.8%show the main content within Google’s 2.5 seconds
- 46.2%pass all three Core Web Vitals with real shoppers
- The median store scored 34 out of 100 on Google’s phone test. 81.2% scored under 50; 0.8% reached 90 or more.
- On a mid-range phone, the median store showed its main content after 11.7 seconds. Google counts 2.5 seconds as good. Only 0.8% of stores made it, and 97.3% took longer than 4 seconds.
- Real shoppers fare better than the lab test, which simulates a mid-range phone on a slow mobile connection: at the 75th percentile of real Chrome visits the median store showed its main content in 2 seconds. Still, only 46.2% of the 353 stores with real-user data pass all three Core Web Vitals. Responsiveness is the weak one: 63% pass on interaction speed.
- The median home page weighs 7.1 MB over 341 requests, including 2,603 KB of JavaScript and 1,617 KB of images.
- The median store runs 20 third-party apps and tags on its home page. The more it runs, the lower it scores: 0-10 apps, median score 49; 11-20 apps, median score 37; 21-30 apps, median score 32; 31+ apps, median score 28.
- The problem Google flags most often: too much work on the main thread, on 91.6% of stores.
Stores by score
Google performance score on a phone, in bands of ten. 367 stores.
The apps on the page.
Reviews, email and SMS pop-ups, chat, analytics and ad tags each load their own code. Google names them in every test, so we counted them. Hosting, CDNs, font services and the store platform itself are not counted as apps.
| Third party | Kind | On how many stores | Share |
|---|---|---|---|
| Google Tag Manager | Tag manager | 74.7% | |
| Google Ads (DoubleClick) | Ads | 63.8% | |
| Klaviyo | Ads | 53.4% | |
| Social | 42% | ||
| Bing Ads | Ads | 41.1% | |
| TikTok | Social | 29.7% | |
| Social | 27.8% | ||
| OneTrust | Cookie banner | 27.2% | |
| Clarity | Utility | 27% | |
| Attentive | Ads | 25.1% | |
| Yotpo | Marketing | 24.3% | |
| Social | 18.8% | ||
| AppNexus | Ads | 18% | |
| Snapchat | Analytics | 17.4% | |
| Impact | Ads | 17.4% |
Look up any of the 34 most common apps
The heaviest, by time the phone spends running them
| Third party | Median main-thread time | Time | Size |
|---|---|---|---|
| AudioEye | 1.28 s | 295 KB | |
| Signifyd | 1.13 s | 258 KB | |
| Forethought | 0.56 s | 926 KB | |
| Google Tag Manager | 0.51 s | 526 KB | |
| Osano | 0.49 s | 80 KB | |
| Alia | 0.45 s | 504 KB | |
| clairedefermat.com | 0.37 s | 109 KB | |
| Stripe | 0.28 s | 283 KB | |
| 0.26 s | 213 KB | ||
| ContentSquare | 0.25 s | 173 KB |
More apps, lower scores
| Apps and tags on the page | Stores | Median score | Score | Blocking time |
|---|---|---|---|---|
| 0-10 | 85 | 49 | 459 ms | |
| 11-20 | 102 | 37 | 983 ms | |
| 21-30 | 100 | 32 | 1512 ms | |
| 31+ | 80 | 28 | 2388 ms |
Page weight.
- 7.1 MBmedian home page
- 2,603 KBmedian JavaScript
- 1,617 KBmedian images
- 341median requests
One in four stores ships more than 12.3 MB to a phone before a shopper does anything. On 86.1% of stores the main content Google times is an image, usually the hero banner.
What Google flags most.
| Problem | Share of stores flagged | Stores | Median saving |
|---|---|---|---|
| Too much work on the main thread | 91.6% | 1.4 s | |
| Files that block the first paint | 90.2% | 1.1 s | |
| Scripts that keep the phone busy | 87.5% | 1.6 s | |
| Scripts downloaded but never used | 80.7% | 1.1 s | |
| Files not cached for returning visitors | 72.8% | 1 s | |
| Styling downloaded but never used | 65.7% | 0.4 s | |
| Redirects before the page loads | 60.5% | 0.8 s | |
| Images heavier than needed | 55% | 0.5 s | |
| Slow first response from the server | 49.3% | 0.2 s | |
| Old JavaScript shipped to new phones | 39% | 0.2 s |
By platform and category.
Platform is read from the files each page loads, so a store running headless on top of Shopify counts as Shopify only when it still loads Shopify’s files. Medians are shown for groups of eight stores or more.
| Platform | Stores | Share | Median score | Score | Main content |
|---|---|---|---|---|---|
| Shopify | 239 | 65.1% | 36 | 11.3 s | |
| Custom or other | 88 | 24% | 33 | 12.6 s | |
| Salesforce Commerce Cloud | 28 | 7.6% | 26 | 12.1 s | |
| WordPress | 5 | 1.4% | too few | ||
| Magento | 3 | 0.8% | too few | ||
| BigCommerce | 2 | 0.5% | too few | ||
| WooCommerce | 2 | 0.5% | too few |
| Category | Stores | Median score | Score | Main content | Within 2.5 s |
|---|---|---|---|---|---|
| Pets | 18 | 45 | 8.5 s | 5.6% | |
| Gifts and paper | 23 | 43 | 17.5 s | 0% | |
| Food and drink | 39 | 41 | 13.6 s | 0% | |
| Fitness and wellness | 32 | 38 | 10.3 s | 0% | |
| Tech accessories | 19 | 37 | 11.5 s | 0% | |
| Beauty and personal care | 43 | 35 | 10.4 s | 2.3% | |
| Baby and kids | 19 | 35 | 11.5 s | 0% | |
| Home and kitchen | 56 | 34 | 11.6 s | 1.8% | |
| Apparel | 59 | 32 | 12 s | 0% | |
| Jewelry and watches | 19 | 31 | 14.1 s | 0% | |
| Outdoor | 26 | 31 | 13.1 s | 0% | |
| Footwear | 14 | 30 | 12.7 s | 0% |
How we did it.
- Sample. 418 US brands with their own online store, chosen across twelve categories from apparel to pet food. It is a curated list of known direct-to-consumer brands, not a random sample of all stores. We dropped sites that failed to load, duplicates that land on the same store, and brands based outside the US, which left 367.
- Test. Google PageSpeed Insights API, mobile strategy (Lighthouse lab test on an emulated mid-range phone with a throttled connection), home page of each store, one run per store. Lighthouse 13.5.0. Runs between September 30, 2026 and September 30, 2026.
- Real visitors. Chrome UX Report (real Chrome users, 75th percentile over 28 days) where Google has enough traffic for the page or its origin. 353 of the 367 stores had enough traffic for Google to report it.
- Limits. Lab scores move a few points between runs, and a home page is not a product page or a checkout. Read single numbers as a snapshot, and the medians as the picture.
- No store is named. This page publishes aggregates only. Third-party tools are named because they are products, not the stores that use them.
Cite this.
7IT Solutions. “US Store Speed 2026: how fast US online stores load on a phone.” September 30, 2026. https://7it.co.il/research/us-store-speed-2026/
Quote and reuse it with a link to this page (CC BY 4.0). The aggregates are in a CSV file.