I was about to dive in to replying to a thread about user-centered design on the NGC4LIB mailing list with my view from a ‘vendor’ when a programmer colleague (not a librarian bone in his body) related this tale to me, after a visit to one of our customers.

“I was in the library talking to some OPAC users” – well that’s an excellent start!

“The first one I spoke to, who was pointing at the shelf-mark code on a result said ‘what’s that gobbledegook for?’ I explained it is supposed to help her find the shelf the book is on. ‘Why can’t I have a little map showing me?’ came the response”

“The second simply said ‘why does this thing keep telling me about the things I can’t have!’ and wandered off”

Well that may not be user-centred design but it’s definitely user-centred feedback.

I can cast my mind back more years than I care to calculate and remember my introduction to the library software world. Just like any other industry, and I’ve been involved with a few, I was greeted with the “but our problems are very different to what you will find elsewhere” message – but they all say that. After a couple of years I realised it was true, we ‘are’ different. Firstly we have a stock control system where the customers insist on giving the stuff back most of the time. Secondly, we have enormously rich structured metadata about that stock which seems to preoccupy us more than the stock itself.

So have we vendors added value to what libraries do? Yes.
Have we got it right? No.
Has anyone, vendor or not, got it right? No. - If they had, we would all be to busy copying them to be having these discussions.

So back to the question in hand - User-centred design is important, very important, but first define your user. The problem we have had and still do, as the recent flurry of activity on this list testifies, is that there are many potential user types for what we historically envisage as a single system.

In an ideal world we would have built the student User Interface; the researcher UI, the high school UI; the reference librarian UI; the enquiry desk UI; the library science UI; the children’s UI; the general public UI; the plugged-in to my citation management software UI; my google-gadget UI – all supported by variations of a catalogue each tuned to deliver results most relevant for each target user group.

Pie in the sky? If you separate the user interface from the underlying cataloguing, indexing and searching capabilities of a solution I think not. If you design the ‘catalogue’ to flexibly support the searching, indexing, and relevance ranking needs of all these user groups you will be able to lay many different [cheap to produce for a library, a vendor, or even a user] UI skins on top of it.

These thin UI skins, on top of a powerful platform, are much simpler and quicker to develop, in days/weeks as against the months/years of our traditional monolithic systems. Because of this we can consider not only user-cantered design, but also user centred development.

In the future I want my programmer colleague to be sat in the corner of a reading room for a few days constructing the user interface that the users want/need with their direct input, help, criticisms, and instant feedback.

This is the premise that is behind much of the development work around the Talis Platform which is proving that this sort of thing is possible. www.talis.com/platform is the place to go to find out more about it. We are looking for people to work with us as we start to roll these services out, but also just as importantly I believe that our experiences in this area will be valuable to the community in general.

(Photo taken by marttj displayed in Flickr)

.

Technorati Tags: Talis, Library, User-centered Design, Catalog, NGC4LIB

Share This