Methodology
How we measure and score global internet connectivity. Our approach is transparent, reproducible, and designed for accuracy.
Overview
The Global Connectivity Index (GCI) measures internet quality from a traveler's perspective. Unlike traditional speed tests that measure peak performance, we focus on real-world usability — what you can expect when working remotely, video calling, or streaming content.
Data Sources
Two-Tier Data System
Baseline Data
Aggregated from public datasets including Ookla's open data, ITU statistics, and regional regulatory reports. Provides coverage for 190+ countries from day one.
Measured Data
Real-world measurements from our community. When a country reaches 200+ measurements per month, it transitions to "Measured" status with higher confidence.
Scoring Formula
The Traveler Score (0-100) combines five metrics:
Metric Definitions
Download Speed
Median download speed in Mbps, measured over 5-second intervals. We use percentile ranking across all countries rather than raw values to account for the wide range of global speeds.
Upload Speed
Median upload speed in Mbps. Critical for video calls and remote work, but weighted lower than download as most activities are download-heavy.
Latency
Round-trip time in milliseconds. Lower is better, so we invert the percentile. Latency affects real-time communication quality more than speed.
Stability Score
Derived from jitter (latency variance) and packet loss estimates. A stable 20 Mbps connection often outperforms an unstable 100 Mbps connection for practical use.
Availability Score
Reflects network coverage and accessibility. Derived from successful connection rates and public infrastructure data.
Confidence Levels
Each country's data comes with a confidence indicator:
Edge vs Origin Measurements
We measure two different connection paths to ensure global fairness:
- Edge Test: Measures connection to the nearest CDN edge node. Represents typical web browsing and streaming experience.
- Origin Test: Measures connection to our origin server in Frankfurt. Useful for comparing raw routing performance.
The Traveler Score primarily uses edge measurements, as they better represent real-world experience.
Anti-Bias Measures
- Median values instead of averages to reduce outlier impact
- Multiple test runs with middle values selected
- Discard measurements from inactive tabs or interrupted tests
- Rate limiting to prevent artificial data manipulation
- Geographic distribution monitoring to ensure representative sampling
Updates
Scores are recalculated daily. Percentile rankings update as new data arrives. The methodology version (currently v1.0) will increment with significant changes.
Data sources & user contributions
In addition to baseline data sources, the Global Connectivity Index incorporates aggregated, anonymous connectivity measurements contributed voluntarily by users worldwide.
These measurements are used exclusively for improving the index and are processed without storing personal data or IP addresses.
Help Improve Our Data
Run a speed test to contribute anonymous measurements from your location.