Streamlined Workflows
A series of initiatives to simplify complex ad campaign workflows, making the UI more intuitive and user-friendly to new campaign managers.
Unless otherwise stated, all work presented on this page is my own.
My Role
Designer
Design Mentor
Timeline
Off-and-on from
2019 – Present
Team Size
3-5 Designers (varied)
1 UX Researcher
Assorted Product Managers
Deliverables
Information Hierarchy Revamp
Workflows
UI Layout Redesign
Background
The company's demand-side ad platform (DSP) was over a decade old.
Whenever new features needed to be added, product managers would cram new inputs into the existing pages, leading to a disorganized, bloated mess. Our forms became a collection of disparate input fields stacked together like a bad game of Tetris.
The DSP's layouts and major workflows were in vast need of an overhaul.
Due to the fact that UI cleanup was not considered revenue-driving, redesign work was rarely officially resourced. Our designers dedicated our "ten percent time" outside of quarterly deliverables to explore and design what a better workflow could look like.
In 2025, some of the improved flows we'd been designing finally got picked up for development.
A sketch representation of our input forms
A Note on My Role
The initiative to streamline workflows was not a single effort, but an umbrella category for work that all designers on our team eventually contributed to.
I advised many of these efforts as a design mentor and subject matter expert for our team, but did not participate in the creation of deliverable assets for some of them.
The work on this page includes only the initiatives I personally created designs for.
Streamlined Targeting
This initiative began as an effort to streamline one of our insights tools called Omniscope. While the revamp of Omniscope as a whole never panned out, the tool shared the same targeting workflow as other parts of the DSP. The targeting workflow was eventually polished in isolation and made it into production.
I planned and participated in much of the early ideation, including idea generation, running brainstorms, diagramming flows, and creating low-fidelity prototypes. The final polished mocks were completed by another designer on the team.
Research
While there was no formal research initiative to kick off this project, we had existing research from product satisfaction surveys and user interviews. As such, we had a robust list of pain points that informed our ideation during the brainstorming phases.
Key Pain Points: The targeting workflow is overwhelming, too many clicks
Crazy 8s
6
TOTAL PARTICIPANTS
1
Product Manager
3
Designers
2
Engineers
The first phase of ideation was a design sprint via Crazy 8s, in which participants quickly sketch as many ideas as possible per round. As this was the start of Covid lockdowns, this step was done remotely to varying degrees of digital artistry.
We completed two rounds individually, one for input flows and one for output flows. Then, we got together to share our concepts and split off to do one more round of sketching. Afterwards, we got together for one big brainstorm to organize our ideas.
My contribution to the activity is below.
Effort x Impact Matrix
After the second round of Crazy 8s, we sorted all of our ideas on an effort x impact matrix.
Low Effort, High Impact
The ideas in this quadrant were passed onto product and engineering to build, with design support wherever necessary.
High Effort, High Impact
Designers would focus their efforts here and dream big.
We also broke down the matrix itself into a flow chart. Each step of the flow chart isolates the matrix breakdown for that specific step of the workflow.
Flow Chart Simplification
Our next task was to take the existing flow and strip it down to a minimal viable set of steps. Using a combination of existing research and usage data from Pendo, we were able to pare down the flow.
The resulting flow served as our product scope to begin designing layouts and prototyping.
The flow charts above and to the right are assets created by the product manager.
They are presented for purely illustrative purposes.
Low-Fidelity Ideation
After working with a product manager to outline our ideal simplified workflow, I and one other designer created a set of low-fidelity prototypes.
We came up with a series of concepts that incorporated ideas from the Crazy 8s brainstorm. In the end we each took a different core concept and built them out into two separate prototypes. Working with our UX researcher, we conducted usability tests to compare and contrast these prototypes.
The prototype I developed is shown below.
In our existing targeting workflow, we displayed all possible targeting options with no way to show or hide items that didn't interest the user.
In this prototype, I explore the idea of only surfacing targeting options when they become relevant.
Iteration
At this point the other designer took over prototyping while I stayed on in an advisory capacity.
The prototype underwent a few more iterations until it reached the final version below.
Final Designs (Mocks by another designer)
This design simplifies the targeting page to only the most common filters as determined by Pendo usage data. Each targeting section lives in a card in the center column of the targeting page with the left column serving as navigation.
When a user clicks the "More Targeting Options" button, a panel slides open to allow them to add more targeting categories to their page.
Full workflow not shown, but hopefully these two screens showcase the intent of the design.
Streamlined Line Creation
Later on, I had the opportunity to bring this project full circle by redesigning the line creation flow.
This project built upon the same principles as the streamlined targeting initiative above, and so we were able to reuse a lot of our existing research and ideation.
Old Flow (not my design)
This is what the previous version of the page looked like. Even with the blur filter on, one can see that the information hierarchy is difficult to follow and that all the fields seem to have been shoved together into one large block of inputs.
Redesign (my design)
Here's the redesign, where I break the page into sections using the card styling and incorporate the left rail summary and "Add More" pattern from the targeting workflow. We've also moved line targeting directly into the line creation flow, rather than having it exist as a separate step after the line creation step.
Impact
Preliminary research from user testing for the simplified designs netted the following results:
72%
increase in perceived ease of use
66%
decrease in task completion time
4.5/5
for total user satisfaction
“It’s more intuitive and a lot more user-friendly than the current one. I especially like that it lets you select more targeting options if you need to and that you’re not forced to look at everything at once.”
“The navigation seems really easy once you just get to work. I really like the layout of it. It’s quick and easy to figure out what you need and quick and easy to edit.”
Closing Thoughts
There were many other similar initiatives that our design team worked on that I did not touch on in this case study.
In general, these series of initiatives were a very start-stop process. As most of this work was never officially prioritized, we often had to delay them to work on other things. Effective design is timeless, though, and the work we did to improve these existing workflows remained relevant throughout the years until we could finally implement them.
While not all of this work made it to production, it is gratifying to see that none of it went to waste. All the research and ideation for these early concepts eventually formed the groundwork of other design decisions across the platform.
Enter password from resume to view project
© 2025 Eugenia Lee