T3 reflections: Where to watch for the next big advances in RIA technology

This is a great article. In my opinion, the advisor-centric world of software leaves much to be desired where the user interface is concerned.

Developers are still committing major usability gaffes, like tiny fonts, grey fonts on white backgrounds (or worse, white font on grey background), tiny controls, having to click “submt” or “enter” with every data point, mixing “pretty” graphics in with vital data, and the list of violations of the commonly accepted best practices regarding UI design goes on and on.

Google seems to have mastered much of this, but the guys at Google use the software they purvey.

Until the people who make the software begin using the software (and I mean day-in, day-out use) or begin doing time and motion studies and eye-tracking analysis on real, live users, I’m not sure how things will improve. It seems that all UI designers should at least peruse the usability info on Jakob Nielsen’s site at http://www.useit.com/.

Ben, good points. Incidentally, I’ve noticed in talking to many of these tech vendors that a large portion of them started as an advisory shop and eventually just started selling the software they had developed for their practice, so they ought to be fairly in touch with what it’s like to use on a day-to-day basis. That said, they probably don’t use their platforms any more the way Googlers do.

On the eye-tracking point, Ed O’brien of Fidelity told me that they do major usability testing (including eye tracking, and finger tracking for mobile apps), and I suspect that this practice happens fairly commonly among other major custodians and possibly the bigger tech vendors. Still, evidence about how users approach one’s current design seems like it would be insufficient to inspire the new design ideas that can significantly change the workflow efficiency of an app.

Eye tracking, etc is fine but there is little (no?) proven theory about how these kinds of physiological measures relate to behavior or information processing, e.g., we don’t really know what it means or if it, most important, predicts anything that can be measured.