As for your two other suggestions,

Add a close button to the discovered items to keep the canvas under control.
Add a drag and drop bookmark folder so I can keep track of my discoveries.

I’m *really* glad you made those, I’ve been planning a post on that for a while. That is, the need for useful recommendations in usability reviews…more on that soon :)

Artists think that they possess some superhuman insight that entitles them to make other people miserable. They do not. Nor are they particularly benevolent, nor are they particularly smart. The artists of my acquaintance are no smarter or dumber than anyone else, and just as dangerous when given tools.

Whether the product evolves through user centered design or genius design, the audience must be the hub of the process.

If the user comes off looking awkward on their geek date, they certainly won’t bring their next date to your site, and they probably won’t come back.

If Etsy added user testing to their arsenal, they may find that their profits – like their happy customer’s experiences – became “repeat”-able rather than “discover”-able.

Imagine a test where a potential customer is given a $100 and a shopping list. Tell them they can keep whatever is left over after they’ve filled the list, but that they’ll lose a dollar for every minute that they shop. Now watch whether they use the search box or Flash navigation. Watch whether they can use the Flash navigation at all or whether its inaccessible to them.

One of those blue dots holds some beautiful Chinese beads, but I can’t remember which dots I’ve been to this visit, let alone on a previous visit.

Add a close button to the discovered items to keep the canvas under control.
Add a drag and drop bookmark folder so I can keep track of my discoveries.

And test whether either of those ideas adds to the enjoyment of the tools and the site :)

Creative designers shouldn’t be stifled by usability testing, it should confirm and inform their design rationale.

Of course, clever people can do a very good job of designing an interface without user testing. BUT, there are many things that Etsy could improve – their “cute” naming of their features doesn’t always give a clear scent or indication as to the outcome of a click (what is Alchemy, that is the difference between time machine 1 & 2), there is no “title” / “alt” bahaviour on the navigation, would it be more desirable to show more of the sellers other items in the right hand column, given there is a fair amount of whitespace at the bottom of that column.

Usability testing can ALWAYS be user to validate and inform decisions. I see no reason why this should stifle creativity. If anything, usability testing should allow the designer to test waters that they may otherwise be wary of.

Interesting comments… One theme that seems to be coming through is the notion of experimentation in navigation, both by the designer and the user. As Mr Smiley said, it’s probably best suited to a select market. I don’t see this working particularly well on a banking site, a government institution or an information based site, for example.

Wioota mentioned that constraints and feedback can actually inspire rather than stifle creativity, and I completely agree. The title I chose for this post was definitely tongue in cheek, although it’s a comment I frequently hear online and on the job. Constraints, whether they be imposed from the outset or as a result of iterative usability evaluation, can challenge us to solve problems in ways we haven’t had to before. This can lead to great and inspiring work.

Sometimes I wonder if people are afraid of running user testing for fear of having to rework their designs – Do they have enough time? Are they too attached to their design? Do the results of user testing invalidate their expertise? I don’t believe so, but I certainly think people are sometimes fearful of this.

Unfortunately though, it’s often the ridiculously short time frames and resultant pressures that interfere with this process.

With regard to evaluating success, as Firstwater has touched on, benchmarking the effects of product changes throughout the process can be invaluable in justifying future work. As we all know too well, like most things, it usually comes down to time and money…which is why it doesn’t seem to occur as often as it could.

In that sense, the interface is serving a marketing need, and attempting to engage the user.

I would love to see their plans for evaluating the success of this – does it increase browsing? purchases?