Last time in this space, I mentioned an IBM research article that indicated that IT workers can be dramatically more productive when using computer systems with very fast response times. In a study of programmers, IBM found that system response times of 0.3 second more than doubled programmer transaction output compared to response times of 3 seconds. This was contrary to the conventional wisdom, which said that computer system delays are really not so bad because they give information workers time to think about their next task.
The interesting question becomes, why does decreasing system response times have such a beneficial impact on productivity? I have a couple of theories and I thought I'd float them by you. (You may have some ideas of your own, too, and if so, please comment.)
First, I don't think information workers *need* a lot of time to think about their next task in many cases. We often know what we want to do and we know the steps we need to execute to get there. So the benefit of a computer-imposed pause may be illusory in such cases. Second, I think the human mind starts to wander a lot more quickly than we might at first suspect. After a few milliseconds of inactivity, we start to think about other things, if only subconsciously. Then, when the computer system is ready for us again, we have to shift gears back to the task at hand. In system design, they call that "context switching." I'm sure it's just as inimical to productivity in humans as it is to silicon processors. Third, I suspect that people tend to be more productive when they can get into a rhythm. However, you can't get into a rhythm when using an information system that presents randomly varying delays, especially if those delays are long enough for our minds to wander, and if we already have a good mental roadmap of which tasks we need to perform next.
Of course I'm no psychologist and all of the above is speculation on my part. What do you think? Is computer system responsiveness more important than we may have thought?
Glenn Weadock is a longtime instructor for Global Knowledge and teaches Windows 7, Server 2008, and Active Directory. He has recently co-developed with Mark Wilkins two advanced Server 2008 classes in the Microsoft Official Curriculum. Glenn also consults through his Colorado-based company Independent Software, Inc. and is technical director of MarketCoach Investment Education Software LLC.
The Leader in Network Knowledge
Sub-second response matters
I have been in this business a long time, and I first heard of such a study from IBM at least 25 years ago.
I know that when I am working on a task with many steps, especially if I already know what the next prompt is going to be and how I will respond, that even a half-second seems forever. I really appreciate the programs with count-down functions, or even a moving bar to indicate that the system is still functioning. back in the days of dumb terminals, when a whole screenful of characters (1920) had to be sent before it showed anything, and communication lines were at 2400 baud (240 characters per second), subsecond response was nearly impossible, but as terminals (and programmers) got smarter and communication got faster, the only thing that stood in the way was how quick you could do a database lookup and format the output. Of course, today we also need to add pictures and fonts, as well as colors and backgrounds to make it pretty.
But I digress...
Subsecond response is hard to do in web-based environments, but still a worthy goal overall. I definitely agree that there is a significant productivity increase with less stress on the user.
Post new comment