Carousell

Design a simpler and smarter Ads platform

Carousell’s ad suite enables professional sellers to reach more buyers.

I joined the team during its transition from a simple, limited product to a comprehensive suite. With product maturity comes the complexity that requires thoughtful design to ensure simplicity and effectiveness. I led the strategy and executed several improvements over the course of 2 years.

Overall impact

  • X% purchase uplift from simplified ad mgmt  and analytics dashboard
  • XX% ad performance uplift  from better buyer experience
  • Established design consistency and Long lasting UX vision for ad products

Role

Design Lead: Execution + management. Team of 2.

Responsibility

Research, Strategy, Product design, Visual design

Timeline

2022-2025

Context

Carousell is a second-hand marketplace connecting casual buyers and sellers. As the platform grows in user base, professional sellers advertise on Carousell to grow their business.

When I first joined the team, improvements to the ad suite were driven largely by reactive and tactical responses to feedback from Sales, customer support and buyer team. To look beyond isolated problems, I consolidated past research and conducted new ones.

Problems

  • Ad management is manual and complex.

  • Ad performance is unpredictable due to fluctuating demand and competition. At the same time, ads are often seen as at odds with a good buyer experience.

  • Sellers find it hard to track and understand ad data.

New Ads platform: An overview

Solution and impact overview

I worked hands-on and mentored a senior designer to redesign and iterate Ads platform in three parts: simplifying ad purchase, improving ad performance for sellers and creating new, more meaningful ad analytics.

The new experience led to a X% increase in ad purchase and XX% improvement in ad performance.

Simplifying Ad purchase

I simplified ad selection, billing workflow and introduced new and simpler ad products.

Improving ad performance

I resolved a stalemate between buyer growth and monetization team by focusing on a smarter, intent-based ad experience for buyers

New analytics dashboard

I launched a new analytics dashboard with meaningful insights and intuitive data visualization.

01. Simplying ad purchase

I simplified ad options from 6 to 3 options with better IA.

I initiated the concept of outcome-centric ad to simplify ad mechanism. E xecuted in 3 phases.

I streamlined billing workflow to create consistency for sellers and eliminate ops overhead from refund and rerun request.

Before

Each tool had a different billing flow, creating cognitive load for sellers. Unhappy flows that end up in refund and paused scenarios happen often.

After

Move all tools to Pay-as-you-go billing removes refund scenarios. Introduced Auto Top-up to increase liquidity and removes campaign paused scenarios.

02. Improving ad performance

I resolved a stalemate between buyer growth and monetization with a buyer intent <> ad content framework.

Observed a stalemate in collaboration with buyer team, I discussed with my manager and we came up with a framework to align the user goals in both teams.

Monetization: “Let me show more ads!!”
Growth: “You are eroding buyer experience!!”

Stalemate

We challenged that ad has to be disruptive to the user experience.

To make ad performs better, we need to reframe “make ad performs better” to “make buyer experience for ad useful”, by aligning the buyer intent with ad content.

Reframing

Ad content <> Buyer intent framework
Experiment 1: New ad UI with better product clarity
Application examples: Backend ad serving logic

I redesigned the buyer-facing ad UI to make ads engaging and useful for buyers .

I shipped a homescreen navigation change to encourage buyer discovery.

A  Mismatch in supply & demand

From data, I discovered the root cause of low ad performance: Buyers come to Carousell for high-frequent and low value categories, while top ad spends are in low-frequent and high-value categories.

A cluster approach

I worked with my manager to group related categories into clusters. This group helps buyers from high-frequent categories like Furniture to discover low-frequent but high ad spending categories like Home Services.

The navigation change & dedicated cluster page

From a flat list of 32 categories to 6 clusters, each with dedicated landing page.

Experiment 2: introduce new, media-rich ad format

03. New analytics dashboard

Understand sellers’ data needs and existing behaviours to inform IA and new analytics

I mentored a senior designer and shadowed to support her in interviews.

The synthesis along user journey informed the IA of the seller’s insight dashboard.

Needs and Opportunities

  1. More intuitive, less ‘mathematical’

  2. Prioritize actionable over accounting accuracy

  3. Simple, # simplistic

IA informed by JTBDs

I aligned with the working team on the principles for meaningful data visualization. 

Explorations & rejected versions
Final Data visualization modules

Impact

Each iteration drives both revenue and user experience impact hand-in-hand. Cumulatively, we saw: 

  • X% purchase uplift from simplified ad mgmt  and analytics dashboarD
  • XX% ad performance uplift  from better buyer experience
  • Established design consistency and Long lasting UX vision for ad products

Go to next project