I am currently conducting user testing on the e4.com website (which is why I haven’t posted in a few days!), which I really enjoy. There is nothing more valuable than testing with real users. We work from a script that we use to take our participants on a certain journey through the site, testing features and content as we go.

The test that I’m currently working on is designed more to examine the langauge we use on the site. Is the wording of link meaningful, do headers make sense, how do users decide what to click on? We’ve found some really interesting things. The demographic for this site is young, web savvy users. Unsurprisingly, images are the main deciding factor on what they choose to click on.

Be it a video clip, a link to a programme site or a game – it is generally the image that decides it. Long paragraphs of text are ignored, and are generally seen as something that puts them off a page. The challenge is then to write concise text that explains things accurately.

Another interesting thing is that the tone of the language needs to be just right. Too “hip” and “trendy” and it can be seen as condescending, as one participant put it:

It looks like the site is trying to be cool.

Trying to be cool just isn’t cool! This is mentioned in relation to the use of the word “Yay” for positive ratings and games descriptions used on the E4 site. The alternative that was suggested by a number of participants was for a simple, visual star rating system. Users want to be able to scan a page, and judge what to click on by visual cues, be it the image associated with a link or the star rating. This can be a challenge for the visual design, as too many repeated images can be distracting – but they need to be there!

The results that one gets from this testing is invaluable, but what is also really interesting is the comments and insights you get outside of the parameters of your script. I like to allow a little deviation from the formal script to see how users will naturally interact with a site. While it’s very important to the results to get a consistent set of results besed on consistent script questions, the unscripted feedback can be just as useful.

Red RoutesThe idea of “Red Routes” or common paths through a site is a very interesting one, and it’s one that I’ve had interesting conversations with James Page of Webnographer about. It means that before testing you work out the common user journeys through your site and test based on this. You could get these from web stats or previous tests and there is real value to this approach. You test for deviation from these routes and try and work out why this deviation occured.

However, I wonder if it’s as simple as that. While there is definitely common journeys, hypertext means that users can get to content in a variety of different ways, all valid. It’s something that I’ll be looking into more and will hopefully blog about!

[ Apologies is this post is a little disjointed - it was written between 2 user testing sessions and in a hurry! ]