First of all, go into the step of choosing the server. The nearest servers can be located in the same city you live in for optimal performance because the data is not transmitted far. That's why many businesses put their servers in many areas so that users of their services in that area can use them better.

However, the whole Internet is not so close. Many websites you visit have servers located very far away - in cities, even in other countries. Therefore, many times using Speedtestfast gives quite high results but when downloading files from the server is too far, the speed will be much slower.

The difference in server location is the reason why the results often vary between services, such as Ookla's Speedtest.net, Netflix's Fast, Speedtestfast...

By the second step, the tool will open multiple connections and use up all available network resources. If you just checked and used the network, the result will be incorrect. For example, if you just checked and watched Netflix or downloaded heavy files, the results would often be slower than when you didn't watch, download nothing.

Experimental conditions also affect quite a lot of results. A wired network computer (Ethernet) often results in a higher rate than smartphones and tablets using Wi-Fi. Even different devices, whether connected to the same network, can give different results.

How to improve accuracy?

If you want to check the correct speed test to see if it is the same as the ad, check it in the most ideal condition: the computer has Ethernet, select the nearest server and do nothing during the test.

If you want the results to be most realistic, use online tools, connect to a stable Wi-Fi network, choose a server that is not too close, open music or common tasks and then start testing.

Of course, no matter how ideal the test conditions are, you cannot get the most accurate results. See this as a reference tool, satisfying curiosity.