Mondelez International — DataMockups
Redesigning Power BI dashboards to make complex Sell-In and Sell-Out data clearer for business decisions.
I collaborated on the UX redesign of Power BI dashboards for Sell-In and Sell-Out data. I structured the information for three business profiles (executive, tactical, and operational) and improved navigation, hierarchy, and data accessibility, all on a short timeline.
Company
Mondelez International
Industry
Data UX / Consumer goods / Business intelligence
Role
UX Designer — Data UX
Product type
Power BI dashboards / Business intelligence tool
Timeline
May 2025 – Jun 2025
Team
UX, Data/BI, Client Stakeholders, Business Users, Product/Delivery
Methods
Stakeholder Needs Analysis, IA, Dashboard UX, Wireframes, Data Visualization, Accessibility, Client Validation
Tools
Power BI, Figma, FigJam, Claude
Status
Delivered dashboards / Under NDA
Confidentiality
Part of this project is under NDA: sensitive data, KPIs, and dashboard content are blurred or simplified. I'm happy to share more detail in an interview or portfolio review.
Context
Mondelez International needed to improve the UX of the Power BI dashboards used to analyze Sell-In and Sell-Out data. The dashboards had to serve users with three distinct decision-making levels: executive, tactical, and operational.
The timeline was short and the data context complex: I had to understand the information fast, identify what each profile needed, and translate that into clearer dashboards.
Why it mattered
The underlying work was understanding how each business user reads data, what decisions they need to make, and how information architecture could take cognitive load off a data-heavy environment.
My role
I worked as UX Designer with a Data UX focus, within a broader data and BI team.
- Analyzed stakeholder needs based on previous research and client context.
- Structured dashboard experiences for executive, tactical, and operational users.
- Designed the visual interface (hierarchy, typography, and an accessible color palette) to improve data readability.
- Explored wireframes and layout alternatives.
- Validated the proposals with the client and adjusted based on their feedback.
The challenge
The main challenge was turning complex business data into dashboards that were useful across decision-making levels without overwhelming users.
- User challenge: each profile needed a different level of detail in an experience that stayed clear and easy to navigate.
- Business challenge: Sell-In and Sell-Out information had to become easier to read and use for decisions.
- Design challenge: we had to understand a complex data context and deliver useful proposals in very little time.
Problem statement
How might we help business users understand and navigate Sell-In and Sell-Out data, covering the decision-making needs of executive, tactical, and operational levels?
Goals & success criteria
User goals
- Find the right level of information for each role.
- Reduce cognitive load when reading complex data.
Business goals
- Make the dashboards more useful for decisions in each user profile.
- Improve the consistency and readability of Power BI reports.
Design goals
- Define dashboard structures by decision-making level.
- Improve hierarchy and navigation with accessibility criteria, including color.
- Validate alternatives with the client and iterate.
Process
01 — Discover
Understanding users and the data context
I started by understanding the dashboard context, the type of data involved, and the users who would interact with it.
- Reviewed previous research and stakeholder information.
- Identified the needs of executive, tactical, and operational users.
- Analyzed the differences between Sell-In and Sell-Out data usage.
- Mapped where structure, navigation, and hierarchy needed improvement.
How I used AI
I used Claude as a project workspace: I centralized the relevant information, got up to speed on the data context faster, and identified what to clarify before defining the structure. The final decisions were mine, based on UX criteria, accessibility, and client feedback.
Discovery structure used to connect user profiles with different levels of data detail and decision-making needs.
02 — Define
Translating business needs into dashboard structure
With the users and context clear, I defined dashboard structures for each decision-making level, so information was easier to scan and interpret by role.
- Synthesized stakeholder needs into three user levels: executive, tactical, and operational.
- Defined differentiated structures and prioritized information hierarchy around each role's decisions.
- Aligned the proposed structure with the client before moving into visual exploration.
How I used AI
I used Claude to organize stakeholder inputs and explore ways to structure the dashboard information.
Information architecture used to separate dashboard needs by user level and reduce cognitive load.
03 — Develop
Exploring wireframes for clearer data reading
Once the structure was defined, I explored wireframes and layout alternatives to improve how users scan information and navigate between views.
- Created wireframes for the different user levels and improved the layout hierarchy.
- Explored navigation patterns between views and accessible color usage.
- Validated the ideas with the client and adjusted based on their feedback.
How I used AI
I used Claude to generate wireframe ideas from the project context; I reviewed, adapted, and validated them with the client before moving forward.
Wireframe exploration used to test dashboard structure before moving into final Power BI implementation.
Color exploration used to support clearer reading and more accessible data interpretation.
04 — Deliver
Delivering clearer dashboards for each decision-making level
The result was three Power BI dashboards tailored to executive, tactical, and operational needs, with better navigation, visual clarity, and accessibility.
- Delivered three dashboard structures, one per decision-making level.
- Improved navigation and hierarchy across Sell-In and Sell-Out views, with an accessible color palette.
- Contributed to the final documentation and design decisions.
How I used AI
I used Claude during final refinement to check that the dashboard logic stayed clear across the three user levels.
Final dashboard overview simplified for portfolio purposes to show structure without exposing sensitive business data.
Before / after
Before
A single dashboard structure tried to serve every profile, and Sell-In and Sell-Out data was hard to scan and interpret.
After
The dashboards ended up organized by decision-making level (executive, tactical, and operational), with better navigation, hierarchy, and accessibility.
Final solution
The proposal focused on three improvements:
Role-based dashboard structure
The dashboards were organized by decision-making level, so each user gets the level of information they need.
Clearer information hierarchy
Data was prioritized and grouped to reduce cognitive load in a very dense experience.
More accessible data visualization
Color and layout decisions improved readability and access to business information.
Impact & results
User value
Business users access and understand Sell-In and Sell-Out data through clearer structures.
Business value
Three dashboards tailored to executive, tactical, and operational decision-making.
Product value
Better navigation, hierarchy, and data accessibility in Power BI.
Team / process value
AI sped up the work in a complex data context, with decisions validated with the client.
Obstacles & trade-offs
The hard balance was between data density, clarity, accessibility, and delivery speed: the information had to be structured for each profile without cutting it down.
- Understanding a complex data context in little time.
- Designing for three user levels with different needs.
- Balancing dashboard density with readability.
- Improving accessibility without losing business meaning in the data.
What I learned
Data UX is about helping people make sense of information faster and with more confidence, not about showing more data.
- I learned to structure dashboards around user roles, not around the available data.
- AI helped me understand a complex domain faster; client validation remained irreplaceable.
- Dashboard UX should support decisions, not just reporting.
- I gained confidence working with Power BI and data visualization.
This is a good snapshot of how I handle a dense, unfamiliar data context on a tight timeline: structure it around what users need, lean on AI to get up to speed faster, and turn a wall of numbers into something people can use.
Stephanie Cacheo — Senior Product Designer / UX Lead