Using design thinking to simplify complex data.


The design challenge!
To create a new B.I dashboard that monitors the manufacturing, finance and sales departments for data analysts to make key decisions that drives each line of business, reduce human errors, provide valuable insights and increase profits.
The Background
MediaTek Inc. is the 4th largest microchip manufacturer in the world that provides chips for wireless communications, televisions, mobile devices and automated driving. In 2014, MediaTek shipped over 1 billion chipsets. With an ever growing need to streamline their business process, MediaTek needed an analytic tool that will allow key stakeholders to see and understand what is happening from all levels of business, manufacturing and finance.
My Role
As the lead designer of a team including a project manager and a junior designer, it was my task to understand the needs of MediaTek and strategize how design can make their business goals into reality. My role was to lead all research and design phases, create the user experience, user interface, develop the visual design language & mentor a junior designer along the way.
Gathering Business Goals
MediaTek is broken into 3 major divisions or business units. These business units control and oversee every aspect of the manufacturing, inventory and the revenue of all the microchips produced and sold in MediaTek. In order to understand how each line of business is performing or under-performing, key stakeholders have to collect data from multiple data hubs and process the data to generate a report that allows a business analyst to understand how that business unit is performing and how MediaTek is impacted. The major pain point with this process is the extensive amount of time lost by users needed to collect data, process data and generate a report. Because data is being processed manually, reports can have errors giving a false overview of the business unit costing MediaTek millions in missed revenue. Below is an example of the type of data being processed and a report being generated.
Strategy
- Create a single source of truth
- Merge MediaTek’s data
- Dashboard specific solution per B.U
- Optimized workflows
Market Dynamics
- Fluctuating supply & demand
- Operational cost
- International shipping & taxation
- Raw material sourcing
Quantitative Research Study
Before visiting MediaTek, I worked with their business operations, analysts and I.T. managers to understand how users are creating reports and the type of data needed to run an analysis. Working with these stakeholders allowed me to see the types of data used, the length of time it takes to gather data to create a report and understand the key performance matrix needed for each business unit to monitor. Below are a few key indicators needed for each department.
Manufacturing
- Chip production
- Production time
- Quality control
- Materials cost
Finance
- Production cost
- Shipping cost
- Operations cost
- Material forecasting
Sales
- Product sales
- Customer sales
- Industry sales
- Forecasting
- Benchmarking
Qualitative Research: Gathering Needs
The questionnaire was sent to 60 users within Mediatek, each department filled out the questionnaire with details about how they work, why they perform certain actions, the length of time it takes to complete tasks and how they use data to gather insights into measurable performance indicators. The result of questionnaire gave me a complete view and brake-down on how each department used data to track performance and how they visualized the data to gather insights. Below is just a few examples of the questionnaire filled out by end-users.
Knowing Our Users
The 3 main stakeholders held a wealth of information about the day-to-day business process at Mediatek. From them, I learned that everyday tasks of the typical analyst slowed down by the manual process of basic information. Each stakeholder expresses the need for a tool that will allow everyone to see the same data and the ability to drill into a business unit and see all the relevant data to it.
The user interviews consisted of 15 in-person user interviews conducted in the user’s office to observe how they work. Each interview lasted 45 minutes long, where I asked questions about how they worked, the biggest issues they have with their current process and ultimately what they wanted and needed for themselves. The underlying theme of the interview sessions was the need for a real-time analytical tool that connects all their data sources together where users no longer need to manual process data but instead has a screen with different business units, click into the business units to see the insights of that unit with the ability to filter the data as they need. Below is our main stockholders representing each business unit’s needs and jobs-to-be-done.
Jenny

Jenny is a manufacturing analyst, her main tasks ensure that specific quotas are being met and that products are being produced when they should. She spends 2-4 hours a day processing data from different data sources to understand key insights.
Thomas

Thomas is a sales analyst, his main task is understanding how MediaTek is impacted by the number of chips being sold compared to forecast rates. His time is spent comparing data and forecasting the future sales. He spends 4-5 hours a day processing data and creating reports.
Danny

Danny is a finance analyst, his task is to oversee the production cost of a family of microchips. Danny spends his day looking at the cost of materials. His entire day is spent overlooking the costs of materials in different data sources and creating reports on how the business is affected.
Our Persona, The Data Analyst
The persona here represents the typical user within MediaTek and gives a in-depth view of my user’s behavior patterns, goals, skills, attitudes, problems, and background information. The persona allows me to understand my target audience and humanize my research to create a understandable representation of what my user’s want and need.
User Journey Of A Business Analyst
Below is the user journey of a data analyst based upon 10 user interviews. Each step is broken into 5 different categories. Each category represents a series of tasks needed in their current workflow to build a report. Analysts from each department of MediaTek undergoes these process each day to view and monitor the daily needs of each department.
Step 1
Monitor Performnce
- Checks daily performance
- B.U. meetings
- Meet with leads
- Check past performance
Step 2
Data Gathering Different D.Bs
- Performance
- Specific areas
- Past performance
- Forecasting
Step 3
Data Cleaning
- Clean data manually
- Correct inaccurate records
Step 4
Report Creation
- Copy/past data to run performance
Step 5
Sharing Results
- Process data in excel create data visualizations
The MediaTek Design Workshop
The workshop brought together the entire line of business, I.T., managers and end-users to focus on one common goal. The beginning of the workshop outlined the overview of an intensive user research study consisting of surveys, questionnaires and walk-alongs. The user research survey served as the staring point of the workshop and allowed Mediatek to understand the problem at hand. Once we established the ground work of user pain-points, the workshop was geared towards designing for needs. During the workshop, we focused heavily on what users needed day-to-day. The focus therefore was on the types of data needed and how data can be used to make informative real-time decisions. In the workshop, we found that different units looked at similar indicators and therefore were related to corresponding units. As part of the workshop, we created exercises that allowed Mediatek to layout a new information architecture that all the business units can use to its fullest. Along with understanding the data, the workshop produced journey maps, user-personas, empathy maps, to-be processes and storyboards outlining their ideal new solution and process.


The Pain Points
In the workshop with MediaTek, we identified key issues plaguing both the business and users when creating a report and understanding the performance of each business unit. The major pain points identified are:
Pain points in journey
- No holistic view of performance of the business
- Long lead time to generate a report
- Manually gathering of data
- Cleaning data manually
- Human errors in reports
The Problem Statement
In the workshop, I validated both the user journey and the pain points and I developed a problem statement to focus our design goals.
Problem Statement
Data analysts do not have a tool allowing them to gather real-time insights to view and monitor performance and business needs.
Developing a How Might We Statement
I created a How Might We statement to focus the design efforts into a question to be solved, this turns the problems into opportunities for generative thinking on how to best solve the problem discovered during my user research.
How Might We…
Allow analysts to have a holistic view of business needs and performance to empower others to make important business decisions through reporting?
Designing Around Our Areas Of Opportunities & H.M.W
After reviewing the analyst’s journey, and the pain points as a group we then identified the key opportunities we want to solve for in our design phase.
Opportunity #1
Provide a holistic view of business needs and performance.
Opportunity #2
Allow each business unit to have it’s own unique view.
Opportunity #3
Break large data sets into key performance indicators.
Opportunity #4
Provide detailed analytics of each key performance indicator.
Drawing The Experience Together
During the last 2 days of the workshop, we focused on drawing the new experience. Each person in the workshop had the opportunity to have his and her voice heard. Teams worked both individually on their ideas while other teams created their solution as one team. Towards the end of the workshop, Mediatek produced an entire end-to-end experience in a sketch format.
Linking User Needs To The Experience
Mediatek expressed lots of interest in redesigning the way their users accessed and engaged with data. During my research phase, it became clear that users expressed many pinpoints with how the data was structured within Mediatek. As a designer, it was my duty to understand the needs of the users and understand the painpoints of the data from the users point-of-view. Below are a set of matrixes of the types of data needed from each unit and why they needed it. After having 10 matrix sheets of structured data laid out in front of us that myself and Mediatek understood the holistic view of data everyone needed and why. This is the first time that senior management and end-user saw this holistic view. Below is a few examples of the key indicators needed to track and maintain each department.
The Workshop Outcome
Because the workshop focused on the needs of the user and based on user research, Mediatek was able to understand holistically the issues at hand. By knowing and confronting the pain points head on, the result was a solution that allowed users to use their expertise into sketches. The sketches below represent the workable outcome of the workshop. The sketches serve as a major pivoting point and allowed me as the lead product designer to understand the needs of Mediatek from a different perspective. Below are just a few examples of the user experience drawn in the workshop.
The design phase…where everything comes together!
After the workshop, I collected the sketches from the group and began the design phase. The sketches in the workshop serve as a starting point for creating the final user experience. With MediaTek, I co-created the basic structure of a dashboard as well as, identified the data needed to be in the user experience. For this design phase, I began by wire framing the end-to-end experience, establish a system for the design and used MediaTek’s design language for the final look and feel.
Whiteboarding The User Experience
My first step in my design process is thinking systematically about the entire user experience and I began by whiteboarding the experience. The whiteboard allows me to try ideas with ease and allows me to see the entire use interface as a system and not as individual parts.
In total, there are 27 info tiles that give MediaTek’s users the ability to view and monitor key insights into all business units. The dashboard is the launching point in which users select which area of the business they need and see detailed information about that line of business. Users have the ability to filter, compare and drill into specific areas of data to understand what is happening within the business unit.
Defining The Information Tiles
The information tiles server 2 purposes, allows users to monitor a key matrix of a business unit and allow for deeper insight into that key matrix. Each info tile is unique and is calculated based on certain thresholds and parameters that are specific to that matrix. Some info tiles indicate a performance rate, while others give details of a line of business.
Creating The Dashboard Experience
The dashboard is the area where users are able to quickly see and make decisions about each business unit. Prior to the dashboard users had to collect data from different sources to understand what is happening with a line of business. The dashboard aggregates the data they need into one area for the entire company to see.
Defining Deeper Insights
Once an info tile is clicked, the user is able to see all the relevant data to that tile. The data within this internal page can display charts and graphs about a measured rate or have tabular data about an entire line of shipping supply of a family of microprocessors.
Designing The Wireframes
The solution is made of a dashboard with KPI tiles allowing the user’s within Mediatek to see an overview of performance. The KPIs allows the user’s within Mediatek to see where areas of the business needs to be improved and allow the users to know why certain things are happening. Once a KPI tile is clicked, the user is able to see every detail of that department, which gives users great flexibility to make realtime decisions.
Designs Using MediaTek's Branding
For the final visual design of the user experience, MediaTek express interest in having it as part of their branding. At the time that I engaged with MediaTek, they rebranded themselves as being a new fun and easy going company that makes great products for it customers. The image below is an example of MediaTek’s brand and represents a fun and bright company ready to create awesome products.
MediaTek’s Branding Guidelines
Here is an example of MediaTek’s branding guidelines used in the final creation of the dashboard experience. From this guideline I was able to use the fonts, colors and layout structure to create a dashboard that represents the core value of MediaTek.
Want to learn more about this case study or others projects?
For details on my design process or interested in collaborating together, feel free to reach out.