How Mercado Libre is working on centralizing Its audiences using a Master ID

Executive Summary

Mercado Libre needed a better way to understand who their users really were. With people using multiple devices, accounts, and partner platforms like Google or Disney+, it became increasingly hard to connect the dots — and harder to measure and optimize their advertising strategies. We partnered with them which enabled the creation of a crucial capability: a unified Identity Resolution system designed to gather and link all user signals under a single, secure ML_ID.

About The Company

Mercado Libre is Latin America’s leading e-commerce technology company. Through its primary platforms, MercadoLibre.com and MercadoPago.com, Mercado Ads, Mercado Envios, and Mercado Crédito, it provides solutions to individuals and companies buying, selling, advertising, and paying for goods online. They enable customers in 18 countries to carry out their commerce offering solutions across the entire value chain.

The Challenge

Mercado Libre faced a common but critical challenge in digital advertising: understanding their users across channels and touchpoints — without compromising data privacy.

Users interact with MELI through multiple accounts, devices, and even third-party services. This fragmented identity made it difficult to accurately attribute marketing performance, leading to inefficiencies in budget allocation and campaign measurement.

At the same time, their existing systems couldn’t deliver insights fast enough. They needed a low-latency analytics system capable of handling massive amounts of user data securely and at scale.

To address their challenges, Mercado Libre (MELI) developed a robust Data Management Platform (DMP) focused on improving audience generation, in partnership with Mutt Data. While the existing DMP utilized internal IDs, a new Identity Resolution component was added to extend its capabilities, specifically enabling the platform to work with anonymous users and introducing the ML_ID as a unifying identifier beyond their internal system user identifier, which works only with logged users.

The Identity Resolution solution addresses the limitations of modern cookie-based tracking and user attribution. By implementing the ML_ID, a unified user identifier, we enable Mercado Libre to connect disparate user signals across devices, accounts, and third-party platforms. This approach overcomes the fragmentation caused by cookie restrictions, privacy regulations, and the increasing use of multiple devices. The solution provides a more accurate and complete view of user journeys, leading to improved attribution and efficient advertising.

Furthermore, the solution facilitates collaboration with partners within data clean room environments. Within our Identity environment Mercado Libre can securely share and analyze data with partners without compromising user privacy. This enables cross-platform insights and joint marketing initiatives, leveraging the power of shared data while adhering to privacy regulations. The data cleanroom integration ensures secure data synchronization and user metrics, contributing to enhanced attribution, targeted advertising, and a more connected user experience across partner ecosystems.

The Solution

We helped MELI to design and implement a scalable Identity Resolution architecture centered around the ML_ID — a unified user identifier built to link diverse signals across platforms.

• We designed and deployed an Identity Resolution API for fast and secure lookups.
• We ensured privacy by hashing all personally identifiable information.
• We orchestrated massive-scale data synchronization jobs using Google Cloud BigQuery for Big Data processing and Cloud Spanner for high throughput & low latency requests handling on attributions.
• Secured data integration within data clean room environments by ensuring privacy through hashing personally identifiable information.
• We implemented daily quality checks to guarantee accuracy, uniqueness, and consistency.

The Results

Thanks to the implementation of a centralized ML_ID now MELI can:

• Address potential revenue loss of 15-20% caused by inefficiencies from the lack of a unified ID.
• Mitigate the risk of losing up to 75% of the initial user base, a common issue with cookie-only ID solutions, thereby significantly expanding the addressable identified audience.
• Impact ~150M monthly unique users by providing a unified identification solution.
• Strategically position to leverage the +12% CAGR projected for identity solutions in the digital advertising market.
• Aim for an approximate 30% increase in campaign effectiveness by implementing a unified ID, mirroring benchmarks set by industry solutions like RampID.

Wrapping up

By building a unified Identity Resolution system and a scalable Data Management Platform, Mercado Libre transformed the way it understands and activates its audiences. The implementation of the ML_ID not only improved attribution accuracy, but also unlocked new capabilities for segmentation, personalization, and cross-platform marketing.

This project strengthened MELI’s data foundation for advertising and beyond — enabling smarter decisions, faster insights, and a more connected user experience at scale.

Why Mercado Libre Loves Working with Us

Throughout our history, Mutt Data has conducted various projects to enhance Mercado Libre’s data capabilities with a special focus on MarTech and AdTech use cases. Mercado Libre has also chosen Mutt once and again to conduct projects related to Generative AI, Machine Learning, and Data Stacks. Mercado Libre and Mutt Data enjoy working with each other due to their teams’ expertise and professionalism. So the Paid Online Media Optimizer for Mercado Libre is not our only lap with this crew. Other joint projects include:

GenAds

Mercado Libre, AWS and us joined forces to tackle the challenge of producing high-performing banner ads with generative artificial intelligence (Gen AI), for dozens of thousands of smaller sellers. Using Amazon Bedrock, Amazon S3, and Amazon DynamoDB, we developed GenAds, an innovative solution that automates the creation of compelling product images, streamlining the process for sellers and enhancing their market presence with image-generating AI. Through GenAds, Mercado Libre helped small sellers through image generation for advertising, enabling sellers of all sizes to attract customers, increase click-through rates (CTR) by an average of 25%, and boost sales effectively.

Multi-Touch Attribution

Mercado Libre needed to improve its online attribution system to measure the incremental lift of each ad touchpoint in conversion journeys, overcoming the limitations of traditional multi-touch attribution (MTA) models. Mutt Data collaborated with Mercado Libre to design and implement a new MTA platform using advanced deep-learning models. This platform accounted for channel interactions, time dependencies, user characteristics, and other business variables, enabling more accurate measurements of each ad touchpoint's true impact. The new attribution model allowed for real-time, campaign-level measurement closer to actual lift estimations. This improved the accuracy of marketing spend optimization, enhanced campaign performance, and informed better decision-making by providing insights into the effectiveness of various marketing strategies across customer journeys.

Boosting Mercado Pago's Customer Base

MercadoPago aimed to increase its Total Payment Volume (TPV) by expanding the range of products used by sellers and introducing new services. The challenge involved both identifying effective recommendations and implementing them through the best channels. Mutt Data developed a comprehensive recommendation framework using Machine Learning models to predict sellers’ propensity to use specific services. This system tailored recommendations based on delivery timing, channel interaction, service relevance, and message content. We significantly accelerated the ML model deployment process. The project led to a 4% nominal increase in sellers selecting MercadoPago as their primary platform, growing from 6% to 10%. Additionally, the implementation of monitoring best practices enabled real-time metric tracking through user-friendly dashboards, further optimizing the platform's performance and decision-making.

Mercado Libre’s Bidding System

Mercado Libre’s Product Ads Team needed to enhance their internal bidding system for promoted items. The existing batch process was suboptimal, leading to inefficient billing and revenue loss, especially for lower bidders. Mutt Data developed a real-time bidding system that optimized Click-Through Rate (CTR) and Conversion Rate (CVR) estimations for each item during searches. The system balanced the impact on Gross Merchandise Volume (GMV) while maximizing ad placement efficiency. It resulted in a more than 25% increase in advertising clicks without significantly affecting organic GMV, demonstrating the effectiveness of the real-time bidding system.

Mercado Libre’s choice to work with us on multiple projects is a testament to the trust we’ve built, thanks to our ability to deliver data-driven, measurable improvements across complex marketing and tech landscapes.

Do you want similar results for your e-commerce platform? Schedule a call with us!