The “Dynamic UX” project - Episode 1
Friday night, I was cleaning the house I had one of those moments which resulted in hitting up the blackboard scrawling lots of slightly out of the box thinking. Rather than put together a horrendously long post, I’ve thrown up this video giving a backgrounder on what I’m thinking about, and hoping to put into action….
Pontificating on Dynamic UX from Rowan Hick on Vimeo.
The basic premise for those that don’t watch the video is - the core web experience over the past number of years, really hasn’t changed - however what we do have, is substantially enhanced productivity through the use of many great frameworks and languages. So let’s look at the core web experience, and how can we alter that based on the behaviour of the user, change the layout and content shown to suit the individual user.
Next Episode, mapping out a system for profiling the behaviour of the user. Look out for a github repo soon.
6 Comments, Comment or Ping
Craig Box
Hey Rowan,
Interesting idea, but on first listen, sounds a bit like Microsoft’s personalised menus, which hide things you don’t use often. Net result, when you -do- go to use something, it’s not there.
There is a lot to be said for consistency of UI layout.
However, I look forward to having my doubts assuaged by your prototype!
Jan 21st, 2009
Carsten Nielsen
I agree, it’s not new news to me that we should be showing users stuff that is relevant to them. The hard part is determining what is relevant. To change the user experience drastically, you need to have drastically convincing evidence that the changes will be relevant.
Perhaps the solution is as simple as suggestions to the user; “I’ve noticed you don’t use the Cheat Codes section, would you like to hide it?” Another approach might be to change the layout without permission and then explain what has happened; “The Cheat Codes section has been hidden to simplify your experience. Undo This?”
Anyways, I’m looking forward to seeing how your approach this Rowan. I’ll be sure to keep an eye on you
Jan 21st, 2009
admin
Craig and Carsten .. thanks for your comments !
@Craig - simply hiding content is probably not a good thing, but weighting visually either size or order, should make a big difference.
@Carsten - Yes it’s not easy determining what and how that content should be shown. Even just toying with the idea of determining what content has been read and how to rank unread content .. is pretty interesting in and of itself.
I was awake at some ungodly hour and toying with how to change up the way of ordering content - rather than by simple “chronological” or “most commented”, but “most relevant” which requires a little bit more inventive tagging schemes - well, specifically tagging with weighting.
Jan 22nd, 2009
Adam Wisniewski
OK, here’s a crazy idea… how about a browser extension that tracks and ranks the pages that you visit. So, when you hit a domain (eg http://ign.com/) it knows that you are most likely to go to the PS3 section and the videos section. So it overlays a little bar at the top of the page with links to those.
This would be something like Google Chromes opening page that shows you your most visited sites. Except that it would index stuff per domain. It could even do this without sending any of your surfing history to the cloud… just parse it all locally in the browser.
It could turn into something more social.. you hook it into an online service, connect with friends… get suggestions based on what your friends surfing habits are, etc.. Or, an API could provide some feedback for the site owners to adjust the site to the users.
Jan 22nd, 2009
Pete Forde
On the current project I’m working on, there was originally a stated intention to make the “dashboard” work in a modular, 3-column layout that allowed the user to move and expand elements that were interesting to them, and have it remember. Over the course of development, we realized that the “iGoogle” pattern might be overkill for what we wanted to do, and we scaled back our expectations for this functionality.
I think that what you’re going to come up against is changing the general expectation of how a website behaves is really going to mess with people’s innate need to organize everything according to a system in their heads. Look at how people flip out every time Facebook changes. I still prefer a layout that was disabled some time ago, even though the current structure is the end-result of a few million dollars spent on UX and case studies and iterations.
Like Carsten said, all of this is fine and dandy until something isn’t where people expect it to be. It’s infuriating when something I’m relying on is trying to be too smart for its own good.
And we’ve all dealt with “clever” UI’s that are actually hindering our ability to use a system. A good example of this is http://pitchfork.tv/ which has a ton of excellent, unique content. However, the person who designed it clearly did not test a working prototype before launch because it’s a total mess. What a shame.
Jan 22nd, 2009
admin
Hey guys, so thanks for the feedback. First and foremost I agree that drastically altering the ux while browsing the site will be disconcerting for the user, however I think massaging of the content and changing layout on entry, based on what you know is probably the avenue I’m going to experiment with.
So here’s my thoughts to kick off with ..
1. sorting content based on relevancy to what you’ve already viewed
2. visually highlighting new and demoting previously viewed content
3. sorting navigation based on derived (via referral or content viewing pattern) preference
4. injecting navigational elements based
5. Providing a viewer feedback mechanism and ability to turn off dynamic ux behaviour
Jan 23rd, 2009
Reply to “The “Dynamic UX” project - Episode 1”