The Vital Role of Connectivity in Site Performance

There's no denying the fundamental role speed plays in the success of a website. While every business wants their site to prosper, unfortunately the speed at which a site loads is a finicky variable. While web page owners can't dictate how fast their site loads, they can make efforts to recognize the factors that create a strong connection, and apply those to their site.

Using Chrome User Experience Report to analyze the impact of internet speed on real-world user experience, Dexecure, a performance enhancing website, set out to find how sites responded to 3G vs 4G connectivity, and if those findings were conclusive across the globe.

Their first experiment measured the effects of connectivity using SEB, Site Experience Benchmark. Using Google as their control variable, Dexecure tested loading times using both 3G and 4G connection. Their results suggest that "by moving from 3G to 4G connection, the SEB for Google improved by 15%."

Moving to a much larger scale, Dexecure targeted 72 countries to see how nations around the globe were impacted by connectivity. Again, using SEB to measure their results and Google as their domain, the connectivities were tested separately. As suspected, not all connectivity strengths are created equal. Venezuela loaded 30% slower than Sweden on 3G connection, and Nigeria operated 25% slower than Sweden on 4G connection. Dexecure coined the hashtags #My3GisNotYour3G and #My4GisNotYour4G to showcase their findings.

Important Takeaways From Their Studies:

Latency and bandwidth play a very important role on the Internet. Businesses are not in control of their site's loading speed. Being aware of what components impact loading speed are important when constructing a site. 3G and 4G connections vary from country to county. Utilizing monitoring tools such as Calibre and Speedcurve can help you better understand your user's environment. In an effort to produce organic findings, synthetic and RUM monitoring should always be performed together.