Maital Rozenboim - Executive Dashboard

Executive Dashboard

| MDClone, Medical Data Platform

Creating a feature based on a whim, not a plan

A dashboard enabling executives to manage projects in the ADAMS medical data platform

My Role

UX Designer & Team Leader

The Comany

Medical, B2B, SaaS, Startup

Tools Used

Axure RP, Figma, Excel

The Team

UX Designer, UI Designer

The Process

Empathize

User Intervews Stakeholder Interviews Competitor Analysis

Define

Persona Building Data Visualization Research

Experiment

Wireframe Validation with Users, PM & Dev

The Wow Factor

UI Design

Iterate

Customer Feedback Implementing Conclusions

The Challenge

What are we doing here?

Executives are managing dozens, sometimes hundreds of research projects. Their time is scarce, and their resources require careful management.

User & Stakeholder Interviews

Wait, but, why?

The Product Manager provided me with no pain points to speak of, only a feature: a dashboard. To understand the user & business needs, I conducted interviews with users and internal stakeholders. I asked them how they were currently using our research project management feature, how they were using other project management software to manage their research projects (if at all), and attempted to determine what additional value this feature can give them.

Insight #1

Don’t Make Me Graph

Users were manually creating graphs and metrics from a csv export of the existing project management table, to use in reports to executives. They were frustrated with repetitively performing a lengthy manual process.

Goal: Remove need for manual, frustrating graphing.

Insight #2

Show Me the Money

Users kept a summary of meaningful everyday metrics like the number of current projects, team members, ROI, and more in their files. They would sum up the numbers manually.

Goal: Show users the metrics they need for everyday work in a front & center manner.

Insight #3

People Have Options

All other project management software that our users were using had dashboard options, and our product looked lacking in comparison.

Goal: Add product value by giving what all our competitors are giving.

Insight #4

Users Have Ideas, Too

One of the executive users illustrated his perfect vision of a visualization, shown below. Each line represents a single project: the X & Y axes show the project phases (progression) and project priority index, the color of the line represents the project’s impact level (high/medium/low) and the dotted line represents new projects.

Goal: Provide the user with the data he worked so hard to show he needs.

User suggested

Persona Building

Who needs this?

I presented user personas to the team, to summarise the research in an easily digestible manner.

Shir, Administrator

45, Data Project Manager

ADAMS Center Administrator

Engagement Level

Usage Type

Entering and Editing Data

Pain Point

Time is wasted on manually creating executive summaries

Need

Full information regarding research projects

What good is the platform if I have to open Excel every day anyway?

Proficiencies

Technical

8/10

Medical

5/10

Data

10/10

Goals

Minimize bottlenecks

Advance valuable research

Raise researcher autonomy

Refael, Executive

60, Medical Doctor

CEO/CFO/Other C-Suite

Engagement Level

Usage Type

Viewing Only

Pain Point

Too busy to go through the minutiae of the data

Need

Birds’ eye view about research projects

I want to see where we stand, not analyse data.

Proficiencies

Technical

5/10

Medical

9/10

Data

7/10

Goals

Increase profits

Improve patient satisfaction

Improve quality of care

Exploring the data & visualization options

What can we show?

At first, the product management team had a vision:

They planned on creating the table-graph the user had sketched.

I explained why this won’t work:

It’s not scalable

Hard to interpret even at a small scale

No glanceable information

And then, product management insisted we need a single visualization of 4 parameters, as the user envisioned, so I created an Excel mockup of a sunburst graph:

A mockup of
PM’s request

I explained why this won’t work:

It’s not scalable

Hard to interpret even at a small scale

Almost no glanceable information

I got to work, refreshing my knowledge of visualization best practices, such as:

Column

Good for simple comparisons

Good for trends over time

Ineffective with long labels

Pie

Users like it, find it familiar

Shows relation between percentages

Confusing in most circumstances

Funnel

Communicates we’re dealing with a process

Helps identify bottlenecks

Difficult to compare 2 data points

Stacked Column

Shows the contributions of components to the whole

Easy comparison in two dimensions

Increases cognitive load

Finally, I took a deep dive into my sample data sets, and made a suggestion: Splitting the data into three separate graphs, letting the user get bite-sized data points.

A modest funnel, this graph shows the relative proportion of projects in each phase, hinting at their passage between them through time. The user should aspire for the graph to be a triangle.

A simple column chart, showing the distribution of projects between the ‘impact’ indices. After discussions of a combined visualization for them, I chose to discard the ‘priority’ index from the dashboard entirely, as its similarity caused confusion.

A stacked column chart, dividing the projects per phase into groups: newly started the phase (may need onboarding); spending a regular amount of time per phase; and stuck in the current phase for too long (might need help). This simplified the metric of length of time spent in phase in a way that presents the actual relevant insight (as gleaned from interviews) and not just a number.

Validation with Users, PM & Dev

Is it up to snuff?

I presented a preliminary wireframe to users, as well as internal stakeholders from the product management and development groups. The users loved it, but product and development had two major hurdles:

Issue #1

Incompatible Graph Library

To track whether projects were advancing in a regular pace through the funnel of project phases, I had recommended... a funnel graph. The existing graphs library R&D were using did not support a funnel graph. Switching graph libraries had a 50% higher effort estimate.

Solution: The funnel was downgraded to a bar chart. Cons: We lost the inherent sense of a process that a funnel conveys. Pros: The user’s eye can still quickly scan for the desired triangle shape, and the dev effort was doable.

Issue #2

Historical Data isn’t Logged

The design I had recommended included several charts that compare data trends throughout a selected period (e.g. year). However, it became apparent that the necessary data wasn’t being logged anywhere.

Solution: The dashboard was split in two. Part 1 - Only current data Part 2 - Historical trends, to be implemented at a later date along with the collection of such data

Final design

Work with UI Designer

Design System is King

The beauty of working with an expert UI designer, is their ability to make something out of nothing. In this case, to create a brand new component that was missing from the design system: a proper dashboard filter, and not a drop-down menu - which was the best I’d found in the existing system, and looked completely misplaced. The use is the same, but the look & feel are a world apart.

Lessons Learned

What Has This Experience Taught Me?

Lesson #1

My Users Can Draw...

... but that doesn’t mean they’re designers. Sketching their thoughts is just another way for users to express their needs. Using these as a design guideline will only limit the usefulness of the design.

Bottom Line: Design for user needs, not wants.

Lesson #2

Alignment is Crucial...

... not only in the design, but also in R&D capabilities. My vision is most likely grander then what Development can create with the prioritization they are handed, and a crucial part of my role is navigating within this tension.

Bottom Line: Aim for the best, compromise for the possible.

Lesson #3

Usage is the Best Compliment.

Our design partner’s middle management (Sharon persona) stopped using their manual processes. Their executives (Rafael persona) embraced using the module, increasing their commitment to the platform. At the next monthly feedback session we got dozens of requests for new dashboard features - more filters, views, customization options.

Bottom Line: User engagement is key for improvement.