Designing an operations platform for Lovepop's retail and fulfillment system
Building operational systems that connect customer demand with manufacturing capacity at Lovepop.
Context
Lovepop operates a hybrid business combining e-commerce, manufacturing, and fulfillment operations. Each customer order triggers a chain of operational processes across multiple teams.
Marketing campaigns generate customer demand through the commercial website, while internal operational systems manage production, inventory, and fulfillment. Before this project, many of these processes were supported by fragmented tools including spreadsheets and legacy systems, creating operational blind spots between marketing, production, and fulfillment teams.
Production teams often lacked real-time visibility into demand, while marketing campaigns could create demand spikes that operations were not prepared for. Fulfillment teams frequently relied on manual checks to track order status across systems.
Lovepop Global Commerce Infrastructure
From Manufacturing in Vietnam to Customer Delivery in the United States
Demand
Supply Chain
Fulfillment
Lovepop's global infrastructure connects manufacturing operations in Vietnam with multiple sales channels and distribution logistics across the United States. Redesigns focused on improving visibility and coordination in website orders and fulfillment operations.
How the scope changed
This project began as a website and marketing initiative to improve sales conversion. While working on it, I saw that the backend systems supporting production, inventory, and fulfillment would not scale with the growth the company was targeting, not because they were failing today, but because they would become a bottleneck as demand increased.
I proposed expanding the scope beyond the website to unify the underlying operational systems, and the CTO and Head of Operations sponsored the initiative. Rather than changing how the factory and distribution teams actually worked day to day, I rolled out the changes incrementally, starting with small improvements and gradually building toward larger ones, so operations gained better visibility without disrupting the physical process itself.
Unified operations dashboard
Website UX improvements

Key Problems
Fragmented operational visibility
Manufacturing, inventory, and fulfillment teams relied on separate tools to track order status.
Teams frequently had to manually verify production queue status, inventory availability, and fulfillment backlog.
Operational decisions often required checking multiple spreadsheets and internal tools.
Marketing demand and operational capacity were disconnected
As the scope expanded, this surfaced real friction: marketing campaigns generated demand spikes that internal systems didn't always reflect in real time, occasionally causing inventory mismatches and delayed fulfillment. It wasn't the reason the project started, but it was one of the frictions worth fixing along the way.
Website purchase flows introduced friction
Parts of the commercial website purchase flow were unnecessarily complex, increasing friction during product selection and checkout.
Solutions
Operational visibility platform
A centralized dashboard allowed teams to understand the state of production and fulfillment in seconds instead of manually checking multiple systems.
Removed manual coordination overhead
We consolidated operational data into a shared interface, reducing reliance on manual communication between production, warehouse, and fulfillment teams.
Improved purchase experience
UX improvements simplified the purchase journey and helped marketing campaigns convert more effectively into completed orders.
Impact
Conversion lift
Over the following year, the redesigned website rolled out continuously, including a new search engine and vendor, fully implemented over four months. The resulting lift in conversion rate didn't look large at first glance, but the CTO called it "significant," tracing it back to real revenue impact.
Earned trust from operations
Operations and forecasting teams, having experienced the quality-of-life improvement the tech team brought to their daily work, began requesting more improvements themselves. Production teams at the factory and the procurement team followed the same pattern, shifting from groups that had to be convinced to ones actively pulling tech into more of their process.
Lessons
Validate relentlessly, not occasionally
Improvements shipped through a continuous queue of A/B tests, run back-to-back with no idle gap between them. Taking outcome validation seriously meant treating it as an ongoing discipline, not a step done once per feature.
Data has to match the story it's used to tell
It's easy to reach for a number that supports the narrative you already believe. The harder, and more useful, discipline is making sure the data actually says what you're claiming it says.
Reflection
Working as a PM wasn't new to me, but this project asked more of that muscle than most. I was holding trade-offs across marketing, operations, and engineering daily, while working directly with the CEO, CMO, and Head of Operations in weekly reviews.
Most of it wasn't glamorous work. It was keeping a lot of small decisions coherent while the business kept moving. I happen to like that kind of work.