Problem
A major pharmaceutical company came to me with a proposal - in an ongoing effort to expand their UX, a particular microsite of theirs had a terrible bounce rate, and they asked me to find out why, and to fix it.
That, in itself, isn't much of a challenge. The real challenge was I couldn't talk to any users, had limited interaction with the major stakeholders, and had access to only a few data points from Google Analytics.
Well, OK then...
Solution
Taking the analytics that we did have, and a quick persona of some people I knew who were similar to the actual user base, the internal team and I used a Lean UX process to give our best hypothesis as to why the site was failing, and what we could do about it. Most importantly, I also did some testing of the current site with some people I knew who were very similar to the user base that we couldn't speak with.
Their feedback confirmed what we suspected - the visual architecture was terrible, the site was overloaded with useless information, and the cognitive load was forcing users to abandon the site in droves, rather than look for what they wanted.
I started with a user flow, then created wires to test. And after multiple iterations, the team and I found a set we liked, and I ran them by my user group, and the results were much better.
After more internal revisions, we moved quickly to final UI, utilizing a recent rebrand of the company to give the site a professional, sleek look. I also added microinteractions that helped the site have a slick but sophisticated appeal.