I use three-four tools to measure website performance: webpagetest.org for the individual pages, Rigor to understand the patterns and catch the problems over time, per location, Page Speed Insights and GA to measure the results. The head of our IT loves GTMetrix, which shows for some reason half number of the speed results compared to webpagetest, rigor and page speed insights which are all normally pretty consistent in generating objective reports. Why GTMetrix shows a lot better results and how to explain it to our infrastructure folks?
Usually performance differences like that come from the connection type used for the testing. WebPageTest defaults to shaping a 5Mbps cable connection (though you can select anything you’d like). I believe GTMetrix only offers connection throttling for registered accounts. It could be something else but that’s the most common cause.
Thank you, Patrick. I wish they wouldn’t sugarcoat the performance results. People tend to be blind to the true numbers and like the reports that show them “nice” and desirable outputs.