A MarTech Success Story:Transforming Mercado Libre’s Paid Media Operations
.svg)
Mercado Libre is the largest online commerce and payments ecosystem in Latin America. Through a suite of technology solutions including Mercado Pago, Mercado Ads, Mercado Envios, and Mercado Creditos they enable customers in 18 countries to carry out their commerce offering solutions across the entire value chain.
The Challenge
With over $1.2 billion in annual marketing investment across multiple channels, Mercado Libre needed a smarter way to manage media spend at scale. The goal was clear: optimize campaign performance to meet aggressive ROAS and revenue targets while maximizing volume. But the path was complex. The team lacked a system that could accurately forecast marginal returns across hundreds of paid campaigns and dynamically reallocate budgets in response to performance shifts. Manual adjustments weren’t fast or granular enough, and existing tooling struggled to balance efficiency with flexibility. As the largest e-commerce platform in Latin America, they needed a solution that could go beyond static rules and guesswork—one that could think in curves, pace spending across time, and react intelligently. That’s where Muttdata came in, bringing deep MarTech experience and a proven track record in spend optimization to design a system built for real-time allocation and long-term scalability.
The Solution
Muttdata developed and deployed the Paid Online Media Optimizer (POMO), a system designed to dynamically allocate daily marketing spend across campaigns and channels—maximizing ROAS while respecting operational constraints like pacing, efficiency targets, and marginal returns. The optimizer captures performance data, interprets business rules (e.g., minimum ROAS thresholds), and models saturation curves to determine how each additional dollar performs. Based on this analysis, it generates optimal spend recommendations, which are then automatically applied to ad platforms via one-click updates.
The system runs in a continuous feedback loop, adjusting budgets daily while monitoring real-time campaign behavior. If performance deviates from expected outcomes, POMO can trigger alerts or pause campaigns autonomously. Built with robust MLOps and software engineering standards, the platform includes fallback logic co-designed with Mercado Libre’s marketing team, ensuring business continuity even when machine learning outputs are unavailable.
The implementation required deep collaboration with MeLi’s Marketing and Data Engineering teams to align with campaign workflows. POMO was fully integrated into Mercado Libre’s Data Lake, supporting seamless reporting, scenario modeling, and cross-channel insights—allowing the team to move from manual operations to intelligent, high-frequency optimizations that scale.
9.6%
55%
