One of the most important thing when it comes to choosing web hosting is site speed. On Google you can search how crucial is to have good page load, and how it can impact on site performance and conversion rates. So, speeding up your website is important not just for better Google ranking, but also for increasing your profit. For example Amazon confirm that:

A 1-second delay in page load time yields:

  • 11% fewer page views
  • 16% decrease in customer satisfaction
  • 7% loss in conversions

So, how does the speed of a website depend on the Data Center?

We made an experiment on how to check page load from different page speed services.

In this experiment we create five identical html test pages (size of 100KB) and put them on five different hosting providers. The goal is to show the best result(page speed) from different hosting companies/data centers.

Our test pages are:

Hosting #1, Hosting #2, Hosting #3, Hosting #4, Hosting #5

For testing we use three services:

– Mbps (www.mbps.net)

– Host_tracker (www.host-tracker.com)

– SiteSpeed (http://sitespeed.me)

Are you ready to lose 7% or more of revenue from your website?

Below you will see why you should carefully select the hosting location for your audition.

Results:

Test Html Mbps

Average Loading Speed/Loading Time

Host-Tracker

Average Loading Speed/Loading Time

Sitespeed

Average Loading Speed/Loading Time

Hosting #1 472.958 Kb/s / 0.21s 292.63 kB/s / 0.34s 406.20 kb/s / 0.24s
Hosting #2 1100.143 Kb/s / 0.093s 224.03 kB/s / 0.45s 250.71 kb/s  / 0.39s
Hosting #3 1098.009 Kb/s / 0.093s 747.95 kB/s / 0.13s 1472.20 kb/s / 0.067s
Hosting #4 217.907 Kb/s / 0.46s 189.37 kB/s  / 0.54s 274.37 kb/s / 0.36s
Hosting #5 100.753 Kb/s  / 1.01s 82.98 kB/s / 1.24 111.72 kb/s / 0.89s

As we can see the load time for the same page, placed in different locations, can vary within one second. According to Amazon, this difference can cost you 7% of the conversion rate. Everything will be even worse if your site includes many additional objects, such as css, pictures and java-scripts.

There are some issues with the accuracy of the data provided by these services. For example, host-tracker.com uses different agent count for each test, or too high speed for Hosting #2, provided by mbps.net, because they have an agent in the same Data Center. But the point can be clearly seen – the same page, located in different Data Centers will be loaded at different speeds.

Load time may be different for every new testing. For much more precise results, we made our service www.mydc.net. With real world ranking, based on continent, region and global testing. You can check details here https://mydc.net/how-it-works/