Driving design strategy to reduce churn
Background
Jungle Scout’s browser extension equips Amazon sellers with the data and insights needed to succeed on the platform. Its product offering consists of the web-app and the browser extension.
The browser extension helps users conduct product-market research. With over 40K daily active events - it has the highest usage in the product offering.
The project took a month to handoff. This included discovering the problem, understanding the ‘Why’, defining the scope (and goals), and designing the solution.
Stakeholders
Worked cross-functionally with PM, tech lead, CX lead, and CS leadership, aligning with senior stakeholders across Product, Design, and Marketing.
Challenge
The browser extension had a 22% monthly churn rate after the freemium period ended.
We could not identify a specific reason for the high churn rate, but it was clear that users did not find enough value in the current experience.
Goal
Our qualitative goal was to increase value by helping users with a better product research experience.
Our quantitative goal was to increase user interaction on the extension.
Let's break it down
Uncovering key problems
I partnered with customer success to analyze churn data from 2,000+ extension users, then interviewed churned users to understand their pain points. I ran force ranking and preference tests to evaluate the extension’s IA and benchmark it against competitors. A heuristic evaluation using revealed critical usability gaps, rated from Moderate to Unusable.
Through six seller interviews, I uncovered key insights:
Most used the extension in isolation, unaware of the complementary web app.
Core actions like filters, bulk select, and tracking products were hidden or missing.
Labels lacked visual hierarchy, and the layout felt cluttered and hard to scan.
These findings helped align stakeholders and drive leadership buy-in for a redesign.
Solutions I designed
To address our project goals, I designed the following:
A central account dropdown where:
Users can now connect their Amazon and Jungle Scout accounts.
See an overview of their business operations.
Redesigned embed cards:
Allows users to customize the label hierarchy based on user's needs.
Call attention to important actions such as ‘Track Product’, which direct users to fine-tune their product research in the web-app.
Introduced a slider component to initiate an increase in user scans of Amazon pages.
New central account dropdown that now allows Amazon and Jungle accounts to be synced
Redesigned search page embed cards
Redesigned product page embed cards that now call attention to 'Track Product' button
New slider component that launches scan pop-up modal
Launch and learning
The launch hit bumps, here’s what we learned
Rolled out to all 450K users at once.
Learning: Next time: Gradual rollout with controlled exposure
50+ bug reports on day one. We lacked proper QA and didn’t test on Windows, our users' primary OS.
Lesson: Hire QA. Expect breakage. Prepare to triage.
No heads-up to users. They woke up to a completely new experience.
Lesson: Introduce change gradually—familiar beats better if it's sudden.
How we fixed the mess!
Cleaning up the chaos
Right off the bat - I integrated myself with customer support teams. I worked overtime and over the weekends fielding complaints and tickets.
My PM and I tracked incoming tickets, and I personally reached out to our most frustrated users to assure them of fixes in progress.
In just one week, I reached out and interviewed 13 users, giving me a firsthand into their pain points with the new release.
Within a month, we shipped 7 updates addressing 50+ tickets. The team earned a new nickname: the agile machine.
Results
Despite a rocky launch, the wins were clear:
+17% in products tracked via extension → Smoother workflows
+9% in filter usage → Deeper engagement with discovery tools
30K+ daily slider clicks → Strong adoption of core feature
8% of all account syncs from extension → New high-value acquisition channel
Churn dropped 13% → Retention improved through better UX
Let's collaborate
Zaidali Rasool , Rhizhome (C) 2025