Amazon Ads · Sr. Product Manager · 2024 to Present
Rebuilding Amazon's Video Creative Editor for Speed
How a personal Canva subscription, a hackathon weekend, and one engineer became the case for rebuilding Amazon's video creative editor from the ground up in a single quarter.
Faster video creative editing
Creative save rate growth
Increase in creatives launched
Context
Amazon's independent sellers, i.e. time-poor entrepreneurs manage every aspect of their business. When they turn to Amazon Ads for help creating video content, they need tools that are fast, simple, and accurate. They don't have creative agencies. They're often using subscription based tools like Canva in another tab.
Problem
Our video creative editor was slow. Internally, it was labelled "acceptable" because we were benchmarking against other internal tools. Our advertisers were benchmarking against Canva. Without an honest external comparison, there was no urgency to fix it. The latency stayed. So did the drop-off.
Approach
01
Make the problem undeniable
I bought a Canva subscription out of pocket and ran both tools side by side for a week. The result was a frame-by-frame teardown showing 30 to 45 second edit latency versus near-instant. I brought it to leadership tied directly to drop-off data in our funnel. The gap became impossible to ignore.
02
Demo the vision
Rather than pitching a rebuild on figma, I partnered with one engineer at a company hackathon to build a working client-side prototype using a third-party rendering library. We demoed live software to senior leadership and engineering. We placed second. More importantly, we walked away with organisational trust and a clear path forward.
03
Ship it in a quarter
Speed was the priority. I drove the commercial licensing conversation, sold the rebuild internally, and delivered a full client-side editor in a single quarter, replacing a server-side architecture that had been in place for years. Six months later, a VP cited this product by name in a company-wide all-hands as an example of the simplicity Amazon Ads needed to deliver at scale.
Outcomes
- →Edit latency cut from 30–45s to near-instant
- →1,000+ new-to-video advertisers onboarded in a single quarter
- →> 30M ad impressions delivered from creatives edited via new-to-video advertisers
Reflections
Latency had been on the radar for a while, but it kept losing the prioritisation battle until I introduced an external benchmark. Customer research doesn't have to be a survey; sometimes it's just being honest about what your product feels like next to the alternative. A working hackathon prototype in front of the right people did the rest. Live software beats a slide deck every time.
Next case study
Unlocking $100M+ ARR with program-level brand safety
Prime Video / Amazon Ads