Problem
Wolters Kluwer's Ovid is a medical research database for Pharmaceutical companies, Doctors, and Medical Librarians. It's a compendium of findings that go all the way back to the US Civil War. So while it's filled with incredibly valuable data, the system was initially created as a CD-Rom, and it still looked like it - cumbersome, unattractive, and unintuitive. The users were starting to flock to other resources, particularly newer users, and revenue was declining.
Solution
We were a small team - a product owner, a developer, a project manager, and me, the sole design resource. We had offshore developers, too, but they were workng on upgrading the search functionality to Apache Solr, so the core front end team was really the four of us.
I usually start with stakeholder interviews, but the product owner had all the details we needed, including some good quantitative metrics, so I was free to start talking to users to learn their pain points.
The contextual inquiry was informative, as it always is. Surprisingly, I learned how a lot of the older users were actually very happy with the system as it was (!). They had adapted it to their tastes and really liked using it, however clunky. The newer users, however - mostly younger, digitally savvy, and impatient with bad user experience - were rejecting the product outright.
The goal then, was to design a system that kept the current user base happy while upgrading it for newer users.
The key moment was when I was interviewing a younger user, a medical librarian, and she offhandedly said "well...no offense, but Ovid isn't Google."
A light bulb went off for me. It all made sense. Her mental model was completely different from the older users simply based on the technological changes that have happened in her lifetime.
Further interviews revealed the same thing - the newer users expected any search functionality - even something as different as Ovid - to operate like Google when the search is initiated.
I knew what I had to do, and I quickly started sketching some potential solutions. I gathered the team together to excitedly run through my hypotheses, and they agreed it was a good way to approach the problem.
I created a user journey to capture the user's task activity, then started with wireframes. I quickly put ideas together, mostly starting with how users initiated the search, clearing the screen of clutter and reducing cognitive load. But I kept a lot of the current functionality available for older users - it was important to allow users to customiae their experience from the get-go
I did multiple iterations of wires, running them by users as I did so, and began crafting a solution that seemed to be working. I also was creating some prototypes in Bootstrap, and testing widgets that we thought users would respond well to, and they did.
Eventually we created a front end that tested so well with users they kept pestering us to get it implemented! Always a good sign.
Eventually, the new Ovid was put in place, and we got extremely positive feedback from the power user base, the younger user base, and even the internal sales team, who told us they were having an easier time getting new sales again because of the new design.