Response Time Goal
Submitted by Alexander Podelko on Fri, 18/11/2005 - 03:40.
performance testing
I started a small research on what response time should be for applications. I know, the topic is being discussed for a long time – still different sources specify quite different numbers. One source recommends 0.1 sec and specifies 1 second as the highest acceptable time, another specifies 8.6 sec. Probably the best sources I found are Beyond performance testing: How fast is fast enough by Scott Barber and How Fast Is Fast Enough by Peter Sevcik, which make an effort to explain why we have such difference and what performance perception can depend on.
Still many questions are open. I haven’t found any specific research for applications (most are about when users abandon a site – something I don’t care much at the time) or any standard for applications (I understand that there are no industry-wide standards, but I’d expect to find some vertical- or area-wide standards). It is still a big difference between 0.1 second and 3 seconds (according Scott, even users who are accustomed to a subsecond response time on a client/server system are happy with a 3-second response time from a Web-based application).
Still many questions are open. I haven’t found any specific research for applications (most are about when users abandon a site – something I don’t care much at the time) or any standard for applications (I understand that there are no industry-wide standards, but I’d expect to find some vertical- or area-wide standards). It is still a big difference between 0.1 second and 3 seconds (according Scott, even users who are accustomed to a subsecond response time on a client/server system are happy with a 3-second response time from a Web-based application).