Understand how your website performs across different devices, browsers, and operating systems.
Device Overview
Performance by Device Type
Device Type Performance (Last 30 Days)
Load Time LCP INP CLS
Desktop 1.8s 1.9s 120ms 0.04
Mobile 3.2s 2.8s 210ms 0.12
Tablet 2.4s 2.3s 160ms 0.08
Mobile typically shows 40-60% slower performance due to:
- Slower processors
- Cellular networks
- Smaller screens (more scrolling)
Device Distribution
Traffic by Device
Desktop ████████████████████████████████ 58%
Mobile ██████████████████████████ 38%
Tablet ████ 4%
Browser Analysis
Performance by Browser
Browser Performance
Browser Version Load Time LCP Users
Chrome 120 1.9s 2.0s 45%
Safari 17 2.1s 2.2s 22%
Firefox 121 2.0s 2.1s 12%
Edge 120 1.9s 2.0s 11%
Samsung Int. 23 3.5s 3.2s 6%
Other - 2.8s 2.6s 4%
Browser-Specific Issues
| Browser | Common Issues |
|---|---|
| Safari | Different scroll behavior, WebP support varies |
| Firefox | Some CSS differences |
| Samsung | Often older engine, slower hardware |
| IE/Edge Legacy | Limited modern feature support |
Version Distribution
Chrome Versions
120 (latest) ████████████████████████ 62%
119 ████████████ 25%
118 ████ 8%
< 118 ██ 5%
Older versions may have:
- Missing optimizations
- Security issues
- Compatibility problems
Operating System Analysis
Performance by OS
Operating System Performance
OS Load Time LCP Traffic
Windows 11 1.8s 1.9s 35%
macOS 1.7s 1.8s 18%
iOS 2.9s 2.6s 22%
Android 3.4s 3.0s 20%
Windows 10 2.0s 2.1s 4%
Linux 1.6s 1.7s 1%
Mobile OS Comparison
Mobile Performance
Load Time LCP INP CLS
iOS 2.9s 2.6s 180ms 0.08
Android 3.4s 3.0s 240ms 0.14
iOS often performs better due to:
- Consistent hardware
- Safari optimizations
- Faster average devices
Screen Size Analysis
Performance by Viewport
Viewport Width Performance
Width Load Time LCP CLS Traffic
< 480px 3.4s 3.0s 0.14 32%
480-768px 3.0s 2.7s 0.10 12%
768-1024px 2.4s 2.3s 0.06 8%
1024-1440px 1.9s 2.0s 0.04 28%
> 1440px 1.8s 1.9s 0.03 20%
Responsive Issues
Small screens may have:
- More layout shifts (CLS)
- Different content loading
- Touch vs click interactions
Connection Type Analysis
Performance by Network
Connection Type Performance
Connection Load Time TTFB Traffic
WiFi 2.0s 150ms 55%
4G 2.8s 220ms 35%
3G 5.2s 480ms 8%
2G 12.5s 1.2s 1%
Offline - - 1%
Network Conditions
| Connection | Typical Bandwidth | Latency |
|---|---|---|
| WiFi/Cable | 10-100+ Mbps | 20-50ms |
| 4G | 5-50 Mbps | 50-100ms |
| 3G | 1-10 Mbps | 100-300ms |
| 2G | 50-250 Kbps | 300-1000ms |
Optimizing for Slow Networks
- Reduce JavaScript payload
- Optimize images aggressively
- Implement service workers
- Use resource hints
CPU Performance
Performance by Device Power
Device Power Performance
Power Level Load Time INP Traffic
High-end 1.6s 90ms 25%
Mid-range 2.4s 160ms 45%
Low-end 4.2s 340ms 30%
Detecting Slow Devices
Zenovay estimates device power from:
- Device model (when available)
- JavaScript execution time
- Memory constraints
- CPU core count
Geographic Impact
Regional Performance
Performance by Region
Region Load Time TTFB Users
North America 1.9s 120ms 40%
Europe 2.1s 150ms 30%
Asia Pacific 2.8s 280ms 20%
South America 3.2s 350ms 6%
Other 3.5s 400ms 4%
CDN Impact
Compare with/without CDN:
CDN Impact on TTFB
Region Without CDN With CDN Improvement
North America 120ms 95ms 21%
Europe 280ms 110ms 61%
Asia Pacific 450ms 140ms 69%
Segmentation & Filtering
Custom Segments
Create device-specific segments:
- Go to Performance → Segments
- Create segment:
- Name: "Slow Mobile"
- Device: Mobile
- Connection: 3G or slower
- Load Time: > 5s
Compare Segments
Side-by-side comparison:
Desktop vs Mobile Comparison
Metric Desktop Mobile Diff
Load Time 1.8s 3.2s +78%
LCP 1.9s 2.8s +47%
INP 120ms 210ms +75%
CLS 0.04 0.12 +200%
Bounce Rate 32% 48% +50%
Alerting by Device
Device-Specific Alerts
Set alerts for specific conditions:
- Go to Settings → Alerts
- Create alert:
- Name: "Mobile Performance Degraded"
- Condition: Mobile LCP > 3s
- Compare: vs last week
- Notify: Slack, Email
Alert Examples
| Alert | Condition |
|---|---|
| Mobile slowdown | Mobile load time increases 20% |
| Browser issue | Chrome LCP spikes |
| New device type | Unusual device appears |
Optimization Strategies
Mobile-First Optimization
Priority for Mobile
1. Reduce JavaScript (biggest impact)
2. Optimize images for mobile
3. Minimize layout shifts
4. Improve tap target sizes
5. Test on real devices
Device-Specific Serving
<!-- Responsive images -->
<picture>
<source
srcset="image-mobile.webp"
media="(max-width: 768px)"
>
<source
srcset="image-desktop.webp"
media="(min-width: 769px)"
>
<img src="image-fallback.jpg" alt="Description">
</picture>
Progressive Enhancement
// Feature detection
if ('IntersectionObserver' in window) {
// Use modern lazy loading
} else {
// Fall back to simpler approach
}
Reporting
Device Performance Report
Generate device-focused reports:
- Go to Reports → Create Report
- Select "Device Performance"
- Choose metrics:
- Performance by device type
- Browser breakdown
- Network conditions
- Schedule delivery
Executive Summary
Device Performance Summary - January 2025
Key Metrics:
• Mobile load time improved 15% from December
• Safari users showing 8% higher CLS
• 4G performance stable
Recommendations:
• Continue mobile optimization
• Investigate Safari layout issues
• Consider lighter experience for 3G
Best Practices
Testing Strategy
- Real device testing: Use actual devices
- Device lab: Multiple OS/browser combinations
- Throttling: Test slow network conditions
- Synthetic + RUM: Lab and real-user data
Prioritization
Focus on:
- Largest device segment (usually mobile)
- Slowest devices in major segment
- Highest-impact issues
- Quick wins first
Performance Budgets by Device
Performance Budgets
Metric Desktop Mobile
LCP < 2.0s < 2.5s
INP < 150ms < 200ms
CLS < 0.05 < 0.1
Load Time < 2.0s < 3.0s
Troubleshooting
No Device Data
Check:
- Script collecting user agent
- Not being blocked by privacy tools
- Sufficient traffic sample
Unexpected Device Types
Review:
- Bot traffic filtering
- Device detection accuracy
- New device categories
Large Performance Gaps
Investigate:
- Network-dependent resources
- Device-specific code paths
- Third-party impact