Mobile Test
Performance in mobile environments
🌐 172.69.130.198
🔥 5 left
Metrics: Median (repeat visit) · JS runtime errors (first/third‑party/unique) ·
Render width overflow
What is the Mobile Performance Test?
Using Playwright, we simulate real mobile device environments to precisely measure your website’s mobile performance and stability.
📊 Measurement Overview
- Tool: Playwright (headless browser, CPU throttling ×4)
- Runs: 4 total per device → skip 1 warm‑up; use median of 3
- Key metrics:
- Repeat‑visit median load time (ms)
- Long Tasks total — TBT‑like (sum over 50 ms)
- JS runtime errors — grouped by first/third‑party
- Render width overflow — body exceeds viewport width
📱 Representative Test Devices (6)
iOS
- iPhone SE (older · small viewport)
- iPhone 11 (mid‑range · common resolution)
- iPhone 15 Pro (latest · high performance)
Android
- Galaxy S9+ (older)
- Galaxy S20 Ultra (high resolution)
- Pixel 5 (reference Android)
Note: Uses Playwright’s built‑in device profiles; if unavailable, a close substitute is used.
🎭 What is Playwright?
- Developed by Microsoft: a modern web automation tool using real browser engines for accurate measurements.
- Headless execution: runs without UI in the background — stable in server environments.
- CPU throttling: limits CPU (×4) to simulate real mobile constraints.
- Precise metrics: measures JS execution time, errors, and rendering performance.
❓ Why can older devices appear faster?
- Lighter assets served: smaller viewport/resolution may receive lighter images/layouts.
- Uniform CPU throttle: same ×4 throttle applied to all, so resource weight matters more than raw device power.
- Conditional loading differences: ads/widgets/scripts may load differently by UA/media queries/breakpoints.
🎯 Why this test matters
- Perceived mobile rendering: focuses on repeat‑visit median and long tasks to capture JS/layout burden.
- Runtime stability: separates JS errors by first/third‑party to pinpoint ownership of issues.
- Responsive suitability: auto‑detects body overflow beyond viewport to catch missing mobile handling.
- Repeatable: fixed devices/runs/throttle/waiting rules — ideal for regression comparison and goals.
| Grade | Score | Performance criteria |
|---|---|---|
| A+ | 90–100 | Median response time: ≤ 800 ms JS runtime errors: 0 Render width overflow: None |
| A | 80–89 | Median response time: ≤ 1200 ms JS runtime errors: ≤ 1 Render width overflow: None |
| B | 70–79 | Median response time: ≤ 2000 ms JS runtime errors: ≤ 2 Render width overflow: Allowed |
| C | 60–69 | Median response time: ≤ 3000 ms JS runtime errors: ≤ 3 Render width overflow: Frequent |
| D | 50–59 | Median response time: ≤ 4000 ms JS runtime errors: ≤ 5 Render width overflow: Severe |
| F | 0–49 | Below the above criteria |
Summary — 6 devices · 4 runs each (1 warm‑up skipped) · CPU ×4 · Metrics: Median / Long Tasks / JS errors (first/third‑party) / Render width overflow
Older devices reading faster can be normal (lighter assets + uniform throttle). This test aims to continuously track mobile rendering cost and stability under realistic conditions.
No results yet
Run a test to view mobile performance results.
No data yet
Run a test to view the raw JSON data.
Sign in to view test history.
Sign in to manage domains.