URLs unleashed(ish)

I am trying something different in the manner in which I start to think about a potential site restructure this time. I have started to ’design’ a URL scheme for the site based on my best bet on where content should logically fit (logic isn’t something that leaps off the site at the moment!).

What I am trying to create is a situation where I can almost A/B test the existing structure that has evolved over the years (and despite best efforts of my predecessors reflects internal needs more than user wants) versus a logical structure based on research, data and experience. I am not expecting either will escape contact with users unscathed!

I am lucky that we have years worth of Google Analytics data to look at as well as previous user testing where not many of the recommendations were implemented (for budget reasons mainly.). Thanks to this I am not just resorting to my own ’gut feeling’ for a change. I have also been looking closely at other similar organisations to see what lessons I can learn from them but then I am still making a few major leaps.

I remain committed to trying to create a website with nicely designed, human readable and hackable URLs. Our site really isn’t big enough (page wise) for this to cause a problem and I like the idea of trying to get a structure that makes sense when read (at least for the top couple of levels).

One thing I am considering – but am not sure whether it is bad practice or not – is ’messing’ with the folder structure in the URLs in places.

One of the main areas of the site *will* be Research (weirdly we really don’t do this well at all at the moment!) and the way I am considering structuring part of it is as below;


So basically you will have a Research landing page, a link from there will be the Research Areas directory/search page and from there you go to the area of research you want to learn about – in this case Regenerative Medicine.

In my original plan the URL would have been /research/areas/regenerative-medicine. Increasingly though I am thinking that it is instead research/regenerative-medicine and that /areas/ is not needed (despite the fact that is likely how it will be organised in the CMS backend). Does this badly break the ability to ’hack’ the URL? Am I worrying over nothing? The back button will still work, links to the /areas/ page will be prominent as well. In some ways I can’t believe I am worrying about this level of detail at this point but to be honest I am finding this a really useful way of visualising what I am trying to achieve and I want to test this internally and with users maybe even before I start working on the wireframes (or maybe doing it alongside would be better but certainly as early as possible.)

I am grateful to the recent burst of publicity that the JISC funded Linking You Toolkit from Lincoln has received as though it was not directly applicable to did clearly spark something in my head (though that spark did lead to 3 hours of working on a Wednesday evening!)

4 thoughts on “URLs unleashed(ish)

  1. Maybe I’m missing something, but what will be the difference between /research and /research/areas ? ‘areas’ seems superfolous to me, unless you have many other items under there that you didn’t mention in the example.


  2. Matt says:

    Yep there are *alot* of other things for that section…./research/ is likely to have 6 or more sections within it i.e. /ethics/ , /people/ , /discoveries/

Comments are closed.