Changing the user perception around a legacy dashboard

My job was to lead the work to uncover the real problems behind the dashboard and solve them, and my first step was to understand why it had been built the way it was.

The answer was simple: at the time, Magazord had no UX or product designer on the team, and the profit-sharing tied to the project only paid out if it shipped within three months. That told me exactly where the project had gone wrong. It was built just for developers, without involve the users and with a very tight deadline.

• The dashboard is too messy

• A lot of information is missing, and I struggle to find the existent information.

• I can't rely on the dashboard, it doesn't give me information fast and clearly enough to make decisions.

• I'd rather download my whole database and load it into a BI tool, even though it's a painful process.

• The dashboard doesn't give me the right information. I had to pull an advanced report, wait more than 5 minutes, and drop it into my excel files.

• I can't even set the frame time which I want discover information

The first step was to map and organize the dashboard indicators to create a clear and intuitive navigation structure. I conducted a card-sorting workshop with the Product and Design teams to define the most appropriate organization for each indicator.

The second (and very important) step was to establish a consistent time frame across all indicators. This allowed users to select a custom date range, ensuring that every metric on the dashboard updated accordingly and remained consistent throughout the experience.

Finally, I redesigned every indicator to be easily understood at a glance. My primary goal was to reduce the cognitive effort required to interpret the data, enabling users to quickly understand what each metric represented and identify where their attention was needed.

This validation process also helped me identify several opportunities for the product roadmap, including real-time dashboards, AI-powered insights, demographic data, and other advanced analytics.

However, this is where the main challenge emerged. The first version of the dashboard was presented to the final decision-makers, and I led the final alignment meeting to define the development roadmap.

During that discussion, they disagreed with the time-range filtering approach and requested the implementation of several indicators that would not follow the same filtering logic.

I won't lie—it was frustrating. Even after presenting the entire research process, usability findings, and the rationale behind the solution, I still had to compromise on a pattern that had been carefully designed and validated with users.

To minimize the impact, I implemented the requested changes in the least disruptive way possible. I added visual flags and contextual tooltips to clearly indicate which indicators did not follow the selected time range, reducing potential confusion and preserving as much consistency as possible.

The results were outstanding, with a satisfaction score of approximately 94%. This validated the design direction and confirmed that the approach I had taken was effective.

During the rollout, the team decided to keep both the legacy and the new dashboards available. Without any additional promotion or incentives, the new dashboard quickly became the most frequently accessed version by a wide margin.

Through customer interviews, I also discovered that many clients had canceled their subscriptions to third-party BI tools because the new dashboard already provided the insights they needed.