News / Infrastructure, Cloud & Telecom
Cloudflare Releases BEACON: 10,000 Sites Analyzed to Measure Real Web Performance Published on 28 September 2026 by Christ-loisele (2 min read)
The BEACON dataset, based on billions of anonymized measurements, reveals significant gaps between rendering engines and industry sectors. A resource for optimizing user experience, according to Cloudflare.
A Global Benchmark Based on Real Data
Cloudflare has released the final version of BEACON, a dataset compiling billions of performance measurements from real browsers across the 10,000 largest websites. Unlike synthetic tests, this data comes from daily user activity, covering all major engines: WebKit (Safari), Blink (Chrome, Edge, Opera), and others. Updated daily in Google BigQuery, it relies on the standards of the RUM Archive , a community project dedicated to user experience metrics.
Closing the perception gap with objective data is central to our mission of helping build a better Internet
Illustration: Lawing Tech Disparities Revealed by Core Web Vitals
The results show that WebKit, used notably by Safari, generally achieves the best scores on Core Web Vitals (LCP, FID, INP). Yet, in 46 countries, its loading (LCP) or interaction (INP) metrics are at least 10% less performant than those of Blink-based browsers. For example, in Cambodia, WebKit accounts for 17.5% of visits but its LCP is 50% slower than Blink’s, illustrating marked geographic variations. The Government and Politics , Health , and Child Safety sectors stand out with high scores, while advertising, religious, and weather sites show the poorest performance.
Optimizing User Experience: Lessons from BEACON
The analysis reveals that the most significant gains in reducing the perception of slowness do not lie in downloading resources but in their discovery and rendering. BEACON introduces sub-metrics to separate loading phases (LCP) and interaction (INP), enabling finer analysis. For companies using Cloudflare, tools like Smart Hints can mitigate delays related to resource discovery. The dataset also exposes hidden gaps behind medians (P75), particularly in visual stability, where performance varies widely among users.
What This Changes Here
For businesses and government agencies in Benin and West Africa, where internet access remains uneven and infrastructure sometimes limited, BEACON could provide a framework for objectively evaluating the performance of local platforms. Public sectors (government, health) could use it to justify investments in web optimization, while private actors (e-commerce, media) would identify targeted improvement areas. For example, if Beninese sites based on WebKit (such as those using Safari) showed delays similar to those observed in Cambodia, technical teams would have data to prioritize fixes. Finally, regulators or telecom operators could use these indicators to measure the impact of infrastructure on user experience, especially in contexts where latency or connection quality varies significantly.
Sources 
