This is the end of the case study!
Thank you for reading!
Always up for a coffee (no sweetener).
Let’s connect!
Improved dashboard functionality and layouts while considering business constraints and user needs.
Power BI Dashboard Redesign for Insights
Initially, the team planned equal task distribution. As the project evolved, I focused on ideation and design, particularly when facing business constraints and conflicting advice.
I championed user research and balanced conceptual and practical aspects. Currently, I'm creating high-fidelity wireframes for key AI flows and preparing for upcoming usability tests.
Window Nation, a window installation services company, connects customers with third-party contractors. They recently launched a daily business intelligence dashboard to support their marketing team.
What is the product?
What was my role?
Business Strategy
Contextual Inquiry
Competitive Analysis and UX Audits
Product Teardown
Design Workshop
Wireframing + Ideation
Industry Sponsored Project
Aug 2024 - Present
Power BI
Figma
FigJam
Microsoft Suite
Roles & Responsibilities
Context
Tools
Students:
Disha Sikaria / Timothy Chiu /
Rachit Bhayana / Natalie Jarrett
Window Nation:
2 Designers / 1 UX Head /
1 Business Strategy Manager
Team
PROBLEM SPACE
The marketing team struggles to engage fully with their dashboard.
The dashboard may not effectively answer critical questions about marketing channels, customer segments, and budget optimization.
Examples of questions:
"Why did radio do badly in Austin this month?"
"Apply percentage change formula to ATL and LOS for today."
There's a need to improve the team's ability to make data-driven decisions based on dashboard insights. Additionally, the company is exploring the potential integration of AI/ML to enhance the dashboard's functionality and user-friendliness.
*content redacted for NDA
RESEARCH
Users want consistency, control, and improved navigation.
Research questions led to our methods.
1 - Evaluating other BI dashboards and AI models.
How has the introduction of the daily business intelligence (BI) dashboard changed the team’s approach to market analysis?
In what ways can areas within the greater umbrella of AI (e.g. ML) be integrated into this system?
In what ways can predictive modeling be connected to Power BI tool?
Our team analyzed 2 types of products: AI assistants and Dashboarding Software. We wanted to find the highlights of these products and how our dashboard could be at the intersection of them.
Dashboards that are familiar to users are preferred. Any new features should be integrated carefully and gradually to allow users to adjust.
Contextual inquiry was conducted with 5 users from the marketing team.
This activity not only helped us identify key issues with the dashboard, but also helped us familiarize ourselves with marketing terminology and how the dashboard is generally used. Previously, we were struggling to understand how the dashboard is interpreted.
Users are generally satisfied with the dashboard.
Users find it frustrating to switch between dashboards on different tabs.
Key Takeaway 1:
Method 1: Competitive Analysis
Evaluation of existing solutions
Method 2: Contextual Inquiry
Observing users complete daily tasks on dashboard
Method 3: Product Teardown
Breaking down existing product
2 - Users find the dashboard sufficient to do tasks.
Identify what is the first thing user does when they come into work
Activity Flow:
Wait for user to get to a task that uses the dashboard
Observe and probe for how task assists in work, cognitive load, and team interactions
Short semi-structured interview to follow up about details and general opinion of dashboard
AI able to lead the user to a solution without the user having to think about a detailed prompt are the most useful. AI can use multimedia to "show" info.
Key Takeaway 2:
User
Key Takeaway 1:
Key Takeaway 2:
User has multiple dashboards open on different tabs
User has Excel on different tab
Users find it frustrating to switch between Excel and marketing dashboard.
Key Takeaway 3:
Different members of the marketing team use different pages on the dashboards, sometimes with no overlap.
Task flow based on contextual inquiry with User 3 helped us map out the common sequences of navigation and events while using dashboard.
Key Takeaway 4:
Inconsistencies between date range across pages makes it confusing to navigate quickly.
AI Insights are difficult to locate, and badly integrated with dashboard. Only provide information about generic trends in natural language.
Data often lacks context. Some numbers are displayed stand-alone, making it difficult to interpret the true meaning and information behind them.
Users like the UI of the dashboard. It is familiar, and an upgrade from the way they were doing things before. Users only expressed that AI could be an aid when explicitly asked.
Key Takeaway 5:
We broke down each aspect of the dashboard page by page, pointing out the stand-out positive and negative aspects of the dashboard from our perspective, taking into account the contextual inquiry.
3 - The teardown revealed major inconsistencies.
Important and useful filters are difficult to find on the dashboard. They are not intuitive to use along with the visualizations on the dashboard.
MY CONTRIBUTION
I created the structure for this research activity. I prepared our team members for how we would make this consistent and participated in the affinity mapping analysis we would do after contextual inquiry.
IDEATION
Our design workshop led to ideas - but many were infeasible.
We produced 50+ ideas in our workshop.
Concept Testing with Window Nation UX Team.
I led this workshop to help us produce design ideas without boxing ourselves into existing products.
1 - We categorized our design requirements and findings, and brainstormed as many ideas as possible as a group.
2 - We refined these ideas at home - allowing everyone to think through ideas at their own pace.
3 - Concept testing with the Window Nation team, to see which ideas they felt fit well with their needs and business constraints.
Window Nation's UX Team gave us insight on the ideas they wanted to prioritize. Many members of the UX team gravitated towards the same ideas, so it was clear to us that certain ideas were standout.
Major point:
We were strongly considering a customizable dashboard of some sort, to allow users to control the content they were seeing.
We received some great feedback - if users could manipulate the data, they would be able to build their own story from it. The data is presented on the dashboard in a particular way to create consistency, and to tell the most accurate story.
50+ ideas
11 ideas presented as sketches
MY CONTRIBUTION
I created the structure for this ideation process. I pushed us to justify our ideas, since the UX team was not familiar with our research. I ideated 7 out of the 11 ideas that Window Nation chose for us to go forward with.
Refining our concepts.
Implementing all 11 ideas in our timeframe was simply infeasible. Additionally, some ideas contradicted one another. We chose 5 narrowed down ideas and sketched out the exact way we saw these appearing on and integrating with the dashboard.
Confusion! - suddenly, there were contradicting opinions from the Window Nation team.
Feeling super lost.
To validate these sketches, we had another round of concept testing.
This time was very different - we received contradicting opinions from various members of the team. This made it difficult for us to move forward.
How would we choose ideas? We thought of reaching out to users again, but the marketing team were not available. And we were running out of time.
We went back to our research to identify the most needed flows. And, we took the Business Manager's advice, to improve the layouts of the dashboard.
One of the key flows we chose was: "Addressing a major change in data".
How does the AI inform the user of the change? How does the user build trust in the AI? How are assumptions validation in the data?
Drawings and annotations for this sketching activity by me!
BUSINESS MANAGER
Focus on feasibility
Excel plugin will not work with Power BI
Use CoPilot as AI Model
Think about improving layouts? (new idea)
UX HEAD
Excel plugin may not have capabilities that are needed for users - too small
Think about feasibility
AI Model can be CoPilot with extra things
UX DESIGNER
Think further outside the box
Too many comments can make the dashboard unusable and overwhelming
Make the AI more innovative than CoPilot
Key flows connecting to the AI ChatBot.
DESIGN
A start on the wireframes.
We are now in the process of designing the wireframes for the flows we have laid out. Here are some of the wireframes and components so far!
After this stage, we will be doing heuristic analysis with the UX Team, and usability tests with the marketing team (our users). Stay tuned for more!
In Progress!