Improving Messaging Insights

Messaging Insights had been released several months prior to me joining the team; however, the usage rates weren't as high as we wanted them to be. By speaking to customers and listening to feedback from our Customer Success Team, we realized that users had difficulties drilling into the data that they needed. Specifically, the filtering left a lot to be desired -- even simple things such as looking at specific campaigns couldn't be done.

We knew we needed to improve the filtering experience by adding in basic functionality and additional paramters that customers needed in order to better understand the performance of their campaigns.

Our team was able to successfully revamp Messaging Insights and improve the usage of the feature. We also received a lot of positive feedback from customers about how easy it was to drill into the data they needed.

We knew that we needed to support additional filters to satisfy the use cases of our customers; however, our existing filtering design wasn't scalable enough and all filters were hidden behind a single dropdown.

In the new design, we wanted to improve both visibility and scalability. I ended up surfacing all nested filters outright and moving the filtering to the side to create more vertical space for filter selection. I also utilized a sentence logic structure that would support additional modifiers and was a pattern that was already familiar to our customers.

The additional filters requested had a variety of requirements. Some filters were simple, such as only needing the ability to multiselect, while others were more complex, such as needing the ability to view nested items. Our existing design system didn't support these additional cases and a few other designers needed similar functionality so I decided to embark on componentizing these elements.

I ended up doing the following:

• Componentizing existing filters and adding a new Ad Lib Pill variant
• Creating a menu panel component with additional variants

For both, it was important to spec out the microinteractions as well as follow the mental models of our customers when organizing information.