I’m fairly confident everyone has used a speedtest site. They are handy and provide a pretty good idea of performance to your computer. Some of the more popular sites that I have used is www.speedtest.net and www.fast.com
In spite of their convenience, there are always caveats and general things to keep in mind when using these sites. External internet connection and related equipment as well as internet variables will affect your overall performance. For example, you might have a weak Wi-Fi signal or busy router that would impact your performance numbers. This is precisely why many ISP support staff request you use a wired connection into your router or modem when possible.
Unfortunately, this gets more complex when you talk about large corporate environments since many other factors are included in the mix. Items such as; Roaming profile file locations, QOS or traffic shaping, flavor of web browser, antivirus (if the test writes and reads from a file), load balancing, system hardening configuration or applications, and of course the end node hardware limitation (i.e. ssd vs mechanical drive, docking station, network dongles. etc..)
In my opinion, there are 2 basic types of speedtests; one that reads/writes a file and one that runs in memory. Most web based speedtest sites that I have used fall into the first category whereas a tool like iperf/iperf3 (https://iperf.fr/iperf-download.php) fall into the latter.
I did a simple test to illustrate the performance differences between the 2 different speed test methods on different laptop models, different browsers, and configurations. For this test, I used a really cool php script from https://librespeed.org/
As you can see results will definitely vary, so one tip I would provide is to perform 5 tests, drop the high/low and average the three remaining results. Generally speaking, iperf3 had the most consistent results in this environment and with the way these devices are configured.