Designing transparency into Etsy's ad ecosystem
400-seller beta group driving retention and revenue
As Design Lead: I owned programming across the four-week, 400-seller beta (planning, monitoring engagement, synthesis), conceptualized eight of the 14 features (three of which ranked top five most wanted), and led rollout strategy for tools prioritized for the permanent experience.
A Complaint That Opened a Bigger Problem
"Budget-constrained" sellers were a high-value group, making up over 1/3 of Ads revenue yet representing only 8% of budget. They saw strong returns and consistently spent their daily budget early. But we knew our product was falling short of their expectations.
Platform Impact
📊 Seller Satisfaction
✔️ 82% seller approval across ads improvements
✔️ 79% rated my performance tool 'most helpful'
✔️ 47% increase in 28-day retention
💰 Feature Value
✔️ 5 features shipped to the full experience
✔️ 3 of my designs ranked in the top five
✔️ 8% high-value Ads budget unlocked
"Is that number good or bad? I have no idea how to account for this information."
- Top Seller
Sellers wanted guidance; raw data without context was creating fatigue, not decisions.
Sellers wanted guidance; raw data without context was creating fatigue, not decisions.
The goal became building a more consultative experience, one that increased transparency and actually empowered sellers to act.
The goal became building a more consultative experience, one that increased transparency and actually empowered sellers to act.
20-person workshop, nearly 50 ideas
20-person workshop, nearly 50 ideas
To quickly generate a wide range of concepts, I planned a 4-day sprint with key Ads stakeholders. I led with an audit of the Etsy Ads experience to inform our direction. The sprint yielded nearly 50 ideas, which we worked with leads to prioritize down to 14.
To quickly generate a wide range of concepts, I planned a 4-day sprint with key Ads stakeholders. I led with an audit of the Etsy Ads experience to inform our direction. The sprint yielded nearly 50 ideas, which we worked with leads to prioritize down to 14.
Four themes in four weeks with 400 sellers
Four themes in four weeks with 400 sellers
My researcher and I designed what we called a "prototype group", similar to beta testing, but with real-time feedback. Each week introduced a new theme, three to six features, distinct hypotheses, and key metrics. Designs and builds happened in tandem requiring tight coordination across the whole team.
My researcher and I designed what we called a "prototype group", similar to beta testing, but with real-time feedback. Each week introduced a new theme, three to six features, distinct hypotheses, and key metrics. Designs and builds happened in tandem requiring tight coordination across the whole team.
✔️ Daily sentiment tracking
✔️ Biweekly moderated research
Three Features Worth Walking Through
We tested 14 features ranging from goal selection to budget insights to data-clarity. These are 3 of the most effective.
1) Listings by Performance
❌ Sellers didn't understand data
❌ Sellers couldn't take action
Sellers needed a faster way to intuit performance and action. I explored visual approaches to surface what mattered: groups by performance with a one-click option to turn off ads for low performers.
Sellers needed a faster way to intuit performance and action. I explored visual approaches to surface what mattered: groups by performance with a one-click option to turn off ads for low performers.
I aligned with engineering on feasibility and logic for dynamic data and edge cases.
I aligned with engineering on feasibility and logic for dynamic data and edge cases.
✅ Majority of sellers turned of low performers while leaving Ads on for high performers (KPI)
✅ 79% of sellers found this feature very helpful making top-ranked
Before rolling out: Investigate the potential ecosystem impact of bulk disabling at scale.
Before rolling out: Investigate the potential ecosystem impact of bulk disabling at scale.
2) Manage Advertised Listings
❌ Poor bulk edit usability
Listing management was a core feature that many sellers didn't know existed or found difficult to use.
Listing management was a core feature that many sellers didn't know existed or found difficult to use.
I reduced friction by moving listing management a level up to the dashboard, and restructured bulk edit actions.
I reduced friction by moving listing management a level up to the dashboard, and restructured bulk edit actions.
I also accounted for mobile, adjusting layout and filters for smaller screens.
I also accounted for mobile, adjusting layout and filters for smaller screens.
✅ Sellers found updates intuitive
3) Detailed Listing Page
❌ Sellers failed to locate
Sellers shuffled between pages and spreadsheets to piece together granular data, which existed but was buried two clicks deep. In a usability study, five of six sellers couldn't find it.
Sellers shuffled between pages and spreadsheets to piece together granular data, which existed but was buried two clicks deep. In a usability study, five of six sellers couldn't find it.
I modernized the page with current components and revised content hierarchy.
I modernized the page with current components and revised content hierarchy.
✅ 77% found it very helpful
✅ 340 sellers clicked through and modified their search terms
Beyond the beta
Outside the prototype workstream, I made incremental improvements to the ads onboarding flow (redesigning the landing page, refining budget selection) driving measurable increases in onboarding completion and 28-day retention.
✅ 82% Ads seller approval
✅ Increase in onboarding completion and 28-day retention
“I could quickly see which ads were doing well or not, and it was easy to make updates.”
- Top Seller
Outcome
The prototype group format powered quick learnings and a more intuitive experience, proving that direct seller access, when structured well, is a powerful way to de-risk decisions before full rollout.
