Re: [Usability] measuring and improving user productivity

[Allan Day <allanpday gmail com>]

Interesting ideas, Matthew. I agree that it is important for the desktop 
to address both the needs of new and more experienced users, and that 
sometimes the latter get forgotten about. To me, something like GNOME Do 
is a really interesting case in this respect. GNOME Do is a tool for 
people who are familiar with the desktop - and it is an extremely useful 
tool for those users. Very fast, very efficient. I think it's telling 
that it is an extra that must be installed by the user, though. In an 
ideal world, the standard desktop would integrate features like this, 
which would tie into basic application launching, and make themselves 
known/discoverable. Interfaces which combine both GUIs and text 
interaction are very promising in terms of what you're talking about, 
therefore.

Banshee is a another good example. Here, you can search and browse using 
the GUI, but there is also a search box with a basic query language (the 
'artist: ' prefix searches by artist, for example). What would be nice 
is to see the behaviour in the GUI fed back into that search box though. 
If I click on 'Kylie' in the browser, the search box should update to 
display 'artist: Kylie'. This would make the search box functionality 
more discoverable. As users became more familiar with the app, so they 
learn to utilise the more 'advanced' search box.

I'm not sure whether 'productivity' is the best term of what you're 
aiming for. It's a difficult concept to work with, and it might be 
unnecessarily limiting. I think what you are pointing to is certain 
aspects of the user experience and particular kinds of 
adjustment/alignment between desktop and user. Perhaps, instead of 
talking about productivity, we should be talking about familiarity? We 
could then talk about mechanisms which leverage familiarity. We could 
also think about how familiarity is promoted and constructed by the 
interfaces that are used by new (ie. unfamiliar) users. An obvious 
example of this is discoverability. You could also think about how 
desktop and applications promote different habits and patterns of use 
(as well as knowledges), though.

Best,

Allan


> Hello, I've had on my mind for a while the topic of user productivity
> which could be thought of as a subject that falls under usability.
>
> Several times I've composed emails on the subject and with each I tend
> to be too detailed or they turn into a rant (or both) and I don't send
> it. So I'm writing briefly to see if others here are interested in
> discussing this. My thoughts in the matter are this:
>
> The goal of a highly usable desktop environment might be to make you
> more productive. That means when you set your mind to a task that is
> possible to accomplish you should be able to accomplish it with the
> least amount of effort. Especially repetitive tasks.
>
> Optimizing for new user productivity may not be beneficial to
> experienced user productivity and vice versa. For example, many call
> centre applications from days gone by used green screen terminal apps
> that used keyboard short cuts to speed up the experienced users.
>
> Testing for user productivity is also a challenge. In my experience
> the more detailed or lengthy the user test is, the less the results of
> the test reflect real world results. Meaning that testing in the lab
> produced results that were inconsistent with what users were doing in
> the real world.
>
> As a daily GNOME user I often get frustrated because the environment
> seems to be working against me. For example, the focus is often in the
> wrong place in the file chooser dialog and programs like GIMP often
> revert to a default setting each time the application is launched.
> This caused me to start thinking about how a series of disparate
> pieces could be changed to work together to improve productivity. Yet
> I'm completely mystified how to measure and test for productivity.
>
> So here we have it, not too long, not too much of a rant. I'm just
> curious if anyone else is thinking about this and if anyone knows how
> to quantify it so that effort can be made to improve it.
>
>