Nielsen Affinity
Media planners have a tough job: sifting through loads of data sets and relying on intuition to target customers can be hard, and often inaccurate.
Nielsen asked the question:
“How might we discover and present the unique story behind a target audience to better empower media planners?”
Team
Building stories that go beyond pure data and numbers is key to understanding consumers. With Ciera Gills (Cornell Tech MBA), Howard Chen and Jessie Gao (Cornell Tech Computer Engineering), we provided a platform to help media planners develop stories around the personalities and interests of their target audience in order to help drive customer sales and media planner success.
Product Manager: Ciera Gills
Product Designer: Terricka Johnson
Engineers: Jessie Gao and Howard Chen
Process & Prototyping
Our initial prototypes sought to develop a visually intuitive design that retained the information of the Nielsen dashboard. With an iterative prototyping methodology and agile workflow, we scrummed our way through the project with an efficient system for development, feedback, implementation and improvement.
User Testing - Prototype 1
After testing our first prototype interface with media planners and non-media planners we received feedback that our product had only made the users feel more confused even though it looked visually appealing. After gaining a better understanding of a media planner's workflow, we broke the planning process down into 3 phases: Demographic, Finding Attributes and Profile Building.
User Testing - Prototype 2
In our second iteration, we broke down the process of a media planner's role in phases of discovery which helped organize the process.
User 2:
- Understood the organization process
- Liked the search but also wanted the ability to selected
User 1:
- Liked being able to step through the process
- Needed more clarity on profile facts and trends
- Wanted more visual representation of selected attributes
User Testing - Prototype 3
Taking feedback from our final iteration, we focused on improving the relationship between the visuals and the stats surrounding selected attritubes.
"This looks great. The one thing that comes to mind here is how to make this target differ or stand out. In doing such items before, you find out you get to very generic things, that don't really tell a good story. It takes a lot of imagination and digging to find a good story."
Technology
By using clustering, we were able to identify key portions or populations and their associations with the world around them. Technologies used include HTML, CSS, Javascript, D3.JS and Python
Attribute points based on user input
Clustering Attributes for Profile Creation