A Metric of Office Non-Productivity?

Worker Using Actual Productivity-Enhancing Tool Image Source

As another generation of operating system and Office Productivity tools emerges from the global leader, The Curmudgeon asks himself - "How much of an improvement will it be?"

Tricky question. This harks back a little to "Why SMART Objectives Could Just Be Called Metrics." since we need some kind of measure in order to answer the question.

Let's try on some metrics for size.

  • lines of text read? Hmmm. Could be really valuable reading or could be the electronic equivalent of a Harlequin novel. No telling.
  • lines of text written? Fairly straightforward. But running on at the electronic mouth can be counterproductive, when you really need to concentrate on something concise.
  • number of email messages sent? No way of telling if you were adding value and actually being productive.
  • number of YouTube videos watched? OK, that's just being silly.
  • number of files opened? Maybe, but if you're just searching for the right file brute-force-style, you might rack up a high score and be wasting time simultaneously.
  • number of files created/modified? Much less likely to be trivial like email, but still no telling if the editing was actually productive in nature.

It's always difficult coming up with a new metric. But what if we looked at the opposite angle - what if we considered NON-productivity? Can we measure that? Is there anything we can identify as definitely being non-productive?

As it happens (and, gentle reader, you might have guessed from the title), The Curmudgeon has an idea for this: mouse-milliseconds.

The mouse is generally used to activate, operate or navigate. It's a tool for using your office productivity tool - a meta-tool. When you're mousing, you're not reading, and you're not writing. Neither absorbing nor generating content. You're trying to follow a link, switch between windows, operate a menu function, select some content for modification, identify an insertion point, etc. Those workers who are creating freehand artwork would be an exception - but you can generally tell that crew by the tablet-and-stylus set replacing their mouse. One might argue that Visio or PowerPoint or Acrobat needs mousework, but if the software was really good at its job, it would lay out your graphics without needing much human intervention (like LaTeX does, but more on that another time).

Hence, mouse-milliseconds - a measure of NON-productivity. OK, so now what?

Well, to answer the question "How much of an improvement will it be?" we can ask "How large a reduction in mouse-milliseconds will we achieve?" To start with, we should benchmark our office workers, then pilot a group over to the new system and compare their mouse-time. Preferably, this would be done double-blind so that we get an unbiased answer. We should also allow some leeway for training and familiarization time, since getting productive with any new tool takes time - no matter how "Intuitive".

That might be worthy undertaking, but I'm certain absolutely no-one** will bother. The Curmudgeon does not believe anyone really believes that they'll become more productive with new software, let alone be willing to subject the question to a quantifiable test.

However, we can get a view of a maximum improvement by looking at an informal study done a few years ago comparing an all-mouse operating approach to an all-keyboard approach. In the study, they operated Microsoft Word first using only the mouse and then using only keyboard shortcuts, and compared the execution times.


Not surprisingly, the keyboard was found to be 3 times faster than mousing. Or (to rephrase to a useful double-negative) 3 times less non-productive. They estimated that such an improvement would net the user 16 hours a year of doing something more productive.

And that's a decent order-of-magnitude answer for The Curmudgeon. "How much of an improvement will it be?" Well, if the new suite was 3 times as fast (and it won't be), it could save 16 hours a year.

Which is a terrible business case. The change will consume much more than that just in training and familiarization time, let alone licensing, hardware, administration, tuning and incidents.

So don't "upgrade". Spend the resources on compulsory "Mavis Beacon Teaches Typing" sessions instead.