Jogan Health Solutions
Jogan Health is an emergency and health staffing/logisitcs company supporting communities throughout the country.
I was hired to help lead the team's effort to provide better reporting to our clients through intuitive and efficient data visualization services. My work also crossed over into supporting our internal teams.
I was hired to help lead the team's effort to provide better reporting to our clients through intuitive and efficient data visualization services. My work also crossed over into supporting our internal teams.
BACKGROUND
When a client requested dashboards, better data visibility, and data integrity I was tagged in from another team to lead the project.
From a product standpoint, I was solely responsible for completing the assigned task in it's entirety, on a very tight deadline. The project required close partnership with our Operations team regarding the collection and handling of data.
I've organized the case study by role into three sections based on my responsibilities to cover the various hats I wore.
From a product standpoint, I was solely responsible for completing the assigned task in it's entirety, on a very tight deadline. The project required close partnership with our Operations team regarding the collection and handling of data.
I've organized the case study by role into three sections based on my responsibilities to cover the various hats I wore.
Product Ownership
After being handed the reins I owned project direction, communications, and timeline through the end of our contract.
Data Science
I was responsible for our database design, Tableau development, and data collection processes from the field into our databases.
UX, UI, & Product Design
I was responsible for the usability and visual appeal of the end product. The project didn't stop after the initial deployment. Any new requirements, issues, or changes flowed through me.
PRODUCT OWNERSHIP
The blank slate challenge...
The first challenge I faced was that we didn't have a technology stack in place to handle large amounts of data or build reports.
My objective here was to identify technologies that could deliver an enterprise scale solution rapidly.
I approached this problem by gathering internal and client stakeholder requirements and followed up with market research.
After research was complete the CEO and I settled on selecting Tableau as the Business Intelligence (BI) software solution.
After gathering requirements from stakeholders I identified two market leaders that would meet our requirements. I evaluated costs, visualization power, analytic power, ability to scale, and ability to play nice with other software. Options were evaluated prior to and post sales demo.
Key Responsibilities
- Evaluate and deploy a new Business Intelligence ecosystem
- Deploy and manage a data migration plan
- Manage client relationship regarding the data visualization project
- Provide Tableau training and onboarding for users and champion the new tool
Data migration challenge
The greatest challenge definitely involved the data migration. We had 100,000+ physical records that needed to be migrated into a digital database.
The migration included two fundamentally different database structures. Prior to the migration, data consisted of a mix of non-uniform reports where data had already been collated.
This meant that the dashboard had to be developed twice over...
I needed to come up with a technological plan to meet our client's needs starting from a blank slate.
DATA SCIENCE
Key Responsibilities
- Database design
- Create data entry forms
- Write code for moving data off forms into our databases
- Data cleaning, shaping, and manipulation in Tableau Prep
- Dashboard development in Tableau
- Tableau Administration
- Determine nomenclature to be used across reports and dashboards
- Update forms and databases after operational changes
New software? No problem... This project posed a significant challenge, that I successfully rose up and met. Prior to this project, I only had limited experience with Tableau briefly as a viewer.
UX, UI, & PRODUCT DESIGN
Time to tie it all together. Our client needed a flexible, functional, intuitive, and visually appealing dashboard system for their project managers and decision makers.
Design and Development Loop
This was an "agile" process. Once I gathered the initial requirements I launched the first version of the dashboard after two weeks of development. Our goal was for the client to have access to the data as quickly as possible. Updates went live weekly after that unless otherwise requested, and I was able to handle most requests at this cadence without building up a backlog. The one exception being during the data migration when we introduced a new dataset and database.
Three types of change requests:
1. Client requests for UI and content
2. My UX/UI improvements
3. Operational changes impacting data
Three types of change requests:
1. Client requests for UI and content
2. My UX/UI improvements
3. Operational changes impacting data
Design Process
Understand
I needed to understand who I was designing for and what their needs were. Who did the team consist of and what were their roles and responsibilities?
Define
I needed to define the information architecture structure, individual visualization components, interaction patterns, and UI elements.
Design
Time to apply some color, padding, and visual hierarchy to the dashboard pages.
Test
I say test here, but in reality the first real UX test was a live instance. The data/information was solid and that was the primary goal at this stage.
Iterate
I made weekly updates as UI feedback, feature requests, operational changes, and UX audits took place.
UNDERSTAND
I needed to understand who I was designing for and what their needs were. Who did the team consist of and what were their roles and responsibilities?
The department essentially consisted of 3 levels of employees in a pyramid hierarchy. People higher up the pyramid need to see all the data and have the ability to drill down. While the people responsible for smaller chunks of data only really needed to see their chunk.
I decided to build one dashboard for the entire team to use. "One dashboard" is used loosely as the dashboard contained multiple pages.
While Tableau provides great features for keeping track of multiple dashboards, one dashboard meant having only one place to look for users. This was something the team appreciated greatly, especially the leaders.
For users who need to use a specific set of filters on a regular basis, Tableau itself provides a feature where you can save filter settings keeping the users from having to reset filters each visit.
While Tableau provides great features for keeping track of multiple dashboards, one dashboard meant having only one place to look for users. This was something the team appreciated greatly, especially the leaders.
For users who need to use a specific set of filters on a regular basis, Tableau itself provides a feature where you can save filter settings keeping the users from having to reset filters each visit.
DEFINE
I needed to define the information architecture structure, individual visualization components, interaction patterns, and UI elements.
While the information architecture was partly determined by the structure of the teams, data was also grouped logically by the various Demographics, Over Time Visualizations, and Location.
The homepage is a great example of the various visualization component decisions. Such as a line graph for showing data over time and a pie graph for showing data as it relates to the whole. A more complicated visualization decision was the Event Type/Dose table on the weekly reports page. It contained data over time, exact data, and data as part of a whole.(Below)
I included two filter types in the dashboard. The weekly report page contained a date filter that was automatically in week format. The weeks were formatted to their requested reporting and weekly schedule. The primary filter was freely manipulatable.
.png)
DESIGN
Time to apply some color, padding, and visual hierarchy to the dashboard pages.
I used Figma to wireframe (sketch) layouts, and concepts before moving into Tableau to finalize designs. Development challenges meant making some alterations when I got to this stage.
I placed total numbers towards the top left of the screen, filters and table visualizations towards the right edge of the screen, and large visualizations in the center.
.png)
A major design challenge I faced was making the dashboard responsive for various screen sizes (I was designing for web only though).
Tableau is not as flexible as the web so this was actually very tricky. It essentially required limiting smaller tiles' such as the "lists", ability to shrink/grow, and allowing the larger tiles more flexibility. I didn't love my solution completely, but functionally it made all the information available at all times and received no negative feedback.
Tableau is not as flexible as the web so this was actually very tricky. It essentially required limiting smaller tiles' such as the "lists", ability to shrink/grow, and allowing the larger tiles more flexibility. I didn't love my solution completely, but functionally it made all the information available at all times and received no negative feedback.
TEST & ITERATE
I say test here, but in reality the first real UX test was a live instance. The data/information was solid and that was the primary goal at this stage.
I made weekly updates as UI feedback, feature requests, operational changes, and UX audits took place.
Making the data migration in the middle of the project allowed to me to get ahead of some design decisions during the middle of the project. I would draft features or changes and present them in meetings for feedback. My favorite way of testing in this way was A/B testing.
FINAL THOUGHTS
Designing for Business Intelligence is tough. Like many business tools, there are a lot of moving parts, various users, various goals, and a mixture of different sciences to be applied. Requirements were changing constantly due to the nature of work being presented. Ultimately my time/focus was split on this project. I would have liked to have solved more of the "problem" around responsiveness from a UX perspective and added more automation into the data side (resource limited).
That all being said the dashboard successfully solved a huge business problem we were facing, a large client was very happy with my work and my internal team members were happy with the leadership and communication I contributed. I grew as a designer and a leader on this project and it felt great supporting our healthcare leaders as they were working to provide care to all of us.
That all being said the dashboard successfully solved a huge business problem we were facing, a large client was very happy with my work and my internal team members were happy with the leadership and communication I contributed. I grew as a designer and a leader on this project and it felt great supporting our healthcare leaders as they were working to provide care to all of us.
