is a measure of the time delay between a request leaving a client browser and being received by the web server.
It is mainly affected by the physical distance between the client and the server. Even if you have a fast connection, a web application that
requires multiple connections to the server will demonstrate poor response times.
Akamai's latency measurements between cities
More information on the causes of latency.
For example, a web application may require 50 elements to be downloaded to display the home page, totalling 200KB.
If you have a 1MB/s connection, you might expect to download the page in 2 seconds - and with parallel downloading, this might be closer to 1 second.
With 100ms latency, however, there is an extra 5 seconds delay in servicing these 50 requests - and if the site uses HTTPS, this could be tripled to 15 seconds.
Anything over 1 second will appear sluggish and more than a 10 second delay is going to make the end-user lose concentration - and patience.
More information on response times.
Ping the server from the command line and note the average round trip time.
You may need to repeat this as the server may respond unusually slowly to the first ping.