Roles: UX designer, UI designer
Tools: Figma, Protopie
Problem Statement: How might we allow a company to effectively use available data?
I was part of a team tasked with creating a dashboard for mock client Adidas' internal-use that satisfied the three given personas representing the business, product management, and design perspectives.
My team created an internal-use dashboard for mock client Adidas, aimed to satisfy three personas with different needs. We made a dashboard that synthesized one common goal: to create better sneakers. The first page gives a performance overview to understand what customers want (categorized by All, Men, and Women). The second page gives insights on the top shoe prototypes and provides feedback on sustainability, reviews from user testing, and measured ROI. All metrics displayed on the dashboard are actionable and directly related to making better, data-driven, shoes, to show Adidas that "impossible is nothing".
Demo video
Collaborators: Amy Zhuang, Mira Mookerjee, Raajat Gupta, Sarah Chen
October – November 2020
1. There is currently no meaningful data driven interactions between the personas.
2. Personas will face fatigue with an oversaturated dashboard.
3. The type of data currently provided is not actionable and therefore not usable.
(detailed research process)
VP Fashion Sneakers, Adidas, Global
Adidas Director of Fashion Sneaker
Marketing for North America
M.A. from Fashion and Textile Design
As we moved from the research phase to ideation, each team member came up with a couple quick sketches of potential graphs and layouts for the dashboard. My sketches focused on creating a dashboard that would stand out to the information overloaded employee. More than convoluted graphs, jumbled words, and vague numbers, I conjured displays that appealed to a human’s natural senses: an image understood without words, colors that praise you or motivate you to work harder.
Our first iteration was bold, creative, and chock full of ideas. We combined all our ideas to create a 4 tab dashboard with expandable widgets. The problem: simply too much. Our dashboard contained every metric you could imagine; it solved everything yet nothing at all.
We had to narrow down our focus to one goal and develop from that. Since Sos and Kenny were our direct users we began our focus there: what do they want to do? They both care about the actual product, while Ansela cares about how the product affects Adidas’ business model. It was now clear that the one goal our dashboard should resolve is the goal to create better shoes.
With our new pristine goal, we made a clean break from the graphs that did not directly contribute to making a better shoe. We then further refined our selection by walking through scenarios where a persona would look at a graph and be able to take some action from the data. If the data existed as an obsolete accessory, we axed it. For example, “If Kenny sees Nike has the top hashtag, he can go to Twitter to learn more.”; “If Sos sees a low sustainability score, she can draw inspiration from our alternative composition graphs.”
We also refined our visuals at this point, condensing multiple graphs on a single page because our previous iteration had too many interactions it made it difficult to understand how to use the dashboard.
From peer critiques, it was clear our dashboard now had a clear message, but lacked some execution. Our visuals still felt crowded, so to increase readability we merged graphs that overlapped, such as ROI graphs.
We polished our visuals as well by adding more white space between graphs
There is nothing more uninviting than a static screen packed with graphs. To delight the user, engage them, and indicate live data updates, we added motion and animation to our prototype.
Loading animations invite our users in, especially less data-inclined users like Sos. On load animations would also give time for real time data to update and display.
When we first thought of a physical control, we ran into the issue of trying to make “a swiss army knife,” as one critique put it. We wanted our control to do everything: slide, push, rotate. But as the list of control functions grew longer, the intuition of our control shrank smaller.
As we stepped back to consider the goal of the project, we realized we had to confine ourselves to a simpler control and figure out a solution with our dashboard. We again evaluated our vision, which pieces of data are necessary to design a better sneaker? And in what sections should they be grouped, so they make the most sense to sneaker design?
Once we settled on the sections to be split into two—Performance Overview and Prototype Production—it was simple to figure out how the control to fit in.
If I were to continue refining this project I would switch from having a physical control to pure touchscreen because it is more integratabtle. The physical control was a great learning tool as a constraint to push our design thinking and prevent "cop outs" that are easy to fall on with touchscreen interactions. I would also conduct user testing with Adidas employees and come up with a list of tasks to test our dashboard's usability.
This was my first time doing a group UX project and I am so thankful for all my teammates who were equally devoted to creating the best dashboard we could. Our open communication is what fueled our success. This project taught me about how to make data visualizations useful and dynamic to multiple users and that every design decision must have a reason.