Driving video adoption through phased experimentation
Generating revenue and a roadmap through dual-track testing and strategic planning
As Design Lead: I owned the strategy for driving video adoption across Etsy's seller platform. I designed three multivariate experiments across partner team surfaces while facilitating a 24-person discovery workshop to shape the long-term roadmap.
The Belief Gap
With over 100M items on Etsy, a listing video is a quick differentiator, yet less than 15% of active listings have one.
Video Adoption Impact
📊 GMS Gains
✔️ $1.1M+ quarterly GMS (2 of 3 tests)
✔️ 231,000+ videos uploaded
⚡️ Velocity & Execution
✔️ 3 experiments designed across 5 weeks
✔️ Phased approach compounding learnings
🫱🏽🫲🏾 Cross-Team Coordination
✔️ 24 people in 2-hour workshop
✔️ 3 teams coordinated for experiments
🎯 Strategic Output
✔️ ~40 feature ideas generated and grouped
✔️ 90% of roadmap established
We knew listings with video get 2x as many orders ($4.75 GMS per listing). But sellers didn't believe it, with less than 28% of them using videos.
Our goal: drive more uploads while building a longer-term video strategy.
"There are already so many steps to list an item. And I don't believe videos really make a difference."
- Top Seller
Two tracks: Prove it now, build it right
To move efficiently, I ran two parallel tracks: progressive experimentation for immediate GMS impact, and a facilitated discovery to build a durable roadmap.
Track 1: Phased Experimentation
I designed three multivariate experiments targeting different placements, increasingly personalized messaging, and building on previous learnings. Each placement lived outside my team's scope, so I coordinated with three partner teams to surface constraints early and keep iterations moving.
Success Criteria
💭 Hypothesis
We believe sharing the value of video with Sellers will result in more video uploads to their listings.
🏆 Key Results
Min. 400k new videos (0.5% coverage lift)
$1.9M GMS
Placements & Outcomes
1️⃣ Listing form
Complexity: Simple
Approach: Static banner
Timeline: 3 days
Impact: $603k GMS
2️⃣ Shop dashboard
Complexity: Medium
Approach: Personalized data
Timeline: 2 weeks
Impact: $800k GMS
3️⃣ Seller app
Complexity: High
Approach: Motion placement
Timeline: Experiment delayed
Impact: Experiment delayed
Surfaced where sellers create and edit listings. It had the simplest flow but significant dependencies to consider.
In three days I finalized two design variants, aligned with Legal on approved copy, and advocated for an A/B structure to get early signal.
✅ $603k GMS
✅ Rollout version B
This tested a personalized banner in the seller's main hub, targeting high visibility without disrupting business-critical tasks.
I explored thumbnail options and worked with my analyst on dynamic messaging. Category-specific competitive data was tailored to each seller, highlighting their high-performing listings without video.
✅ $800k GMS
✅ Rollout version B
This placement was in the Etsy Seller App (ESA). I did a behavioral audit on native banners as this experiment was our steepest learning curve.
I led multiple sessions with the ESA team to scope capabilities and constraints, optimized copy for mobile, and explored auto-playing video to reduce friction at upload.
This A/B experiment was designed but not run as our team was pivoted to an urgent project.
Track 2: Strategy Discovery
In parallel with experimentation I facilitated a two-hour, 24-person discovery workshop to inform longer-term video strategy.
The workshop included multi-stakeholder involvement. I led with a competitive audit and past research, while participants captured notes which I later affinity-mapped into seven themes.
I assigned participants to one of seven boards (one per theme). Each board had "How might we" prompts, relevant notes, and structured idea cards.
The workshop generated ~40 feature ideas. Team leads and I prioritized these into five themes and eight features.
This informed the majority of the upcoming roadmap in just 3 weeks.
✅ 8 features across 5 themes
✅ 90% of roadmap
Reflections
📈 Phased testing compounds
Starting simple validated messaging before layering in personalization.
🖌️ Coordination as a multiplier
With three partner teams, design-led discussions built trust and kept us moving.
🏃🏾♀️ Trust can drive speed
Velocity was the output of building relationships through clear communication, not shortcuts.