Case Study
Campaign management platform for marketing agencies — replacing spreadsheets and scattered tools with one scalable workflow.
3 months
1 PM · Founding Designer (me) · 3 Developers · Marketing Team
UX Design · Design System (0→1) · End-to-End
Figma, FigJam, Notion
INTRODUCTION
8h → 30min time to set up a single campaign
+23% client satisfaction after launch
10× sales growth following the redesign
CONTEXT
WHAT I HAD
An early product built on Ant Design, campaign data still tracked across spreadsheets, and no dedicated design system.
WHAT I HAD TO FIGURE OUT
Who Pickle serves after its pivot, how the campaign workflow scales, and when to invest in the design system.
THE PROBLEM
Agencies stitched campaign data together across spreadsheets, docs, and disconnected tools — every campaign meant re-entering the same information by hand.
GOAL 01
Cut setup time
Get a campaign running in minutes, not hours.
GOAL 02
Unify scattered data
One system of record instead of five disconnected tools.
GOAL 03
Scale with volume
Stay usable as agencies manage more campaigns at once.
1ST ITERATION
The original kanban board made status visible, but it collapsed under real campaign volume. Agencies needed to scan fields across many campaigns at once — not open cards one by one.
01 Initial Idea
The Kanban board forced users to strictly follow a fixed process
02 First Iteration
Table view offered more flexibility & lower learning curve
03 Final Solution
Dynamic table that is flexible to various workflow
FINAL SOLUTION
CONFIGURABLE COLUMNS AND VIEWS
Agencies see the fields that matter to their workflow, without the interface dictating what they can track.
BUILT FOR VOLUME
The experience scales cleanly whether an agency is tracking five campaigns or fifty.
IMPACT
8h → 30min
Time to set up a single campaign, from manual entry to a configured workflow.
+23%
Client satisfaction increase after launch.
10+
Distinct workflows unified into one system.
10×
Sales growth following the redesign.
WHAT I’D DO DIFFERENTLY
The product began on Ant Design and migrated to Untitled UI mid-project. The switch would have been cheaper before components were already built out. Next time, I’d advocate for a custom system from the start — a design system isn’t overhead to add later; it’s how teams avoid rebuilding.
