arrow_back Back Next →
Platform Dashboard

ML Seller Center

A unified seller management platform for Mercado Libre consolidating listings, orders, analytics, and customer communications into a single command center for the 2.5 million active sellers on the platform.

Role

Lead Product Designer

Year

2024

Duration

20 weeks

Deliverables

UI Design, Components, Handoff

Tools

Figma, FigJam

ML Seller Center

ML Seller Center unified dashboard for listings, orders, and performance analytics

01 Overview

Mercado Libre has 2.5 million active sellers across Latin America from individual sellers with 10 listings to enterprise retailers managing catalogs of 50,000+ SKUs. All of them used the same seller interface, which had been designed for the smallest seller and scaled awkwardly to serve the largest.

The ML Seller Center redesign was a platform project: building the information architecture, navigation system, and core screens for a seller tool that works across all seller tiers, with the complexity of the interface scaling appropriately to the complexity of the seller's operation.

02 The Problem

Seller interviews across three countries Brazil, Argentina, and Mexico with sellers at different tiers revealed a split reality. SMB sellers (under 500 active listings) found the platform overwhelming they opened it daily to check orders and then left, never discovering features that would have helped them grow. Enterprise sellers found it underpowered they had built spreadsheet workflows to supplement what the platform couldn't do natively.

The specific failures: no seller could see their order status, inventory levels, and performance metrics on the same screen. Product listing creation required navigating 8 separate form pages. Customer messages and dispute management lived in a completely different interface that didn't share any visual language with the seller dashboard.

Seller research

Seller interviews SMB and enterprise sellers across Brazil, Argentina, and Mexico

03 Process

The project began with a seller tier analysis: defining three operational profiles Solo Seller (1 100 listings), Growing Seller (100 5,000 listings), and Power Seller (5,000+ listings) and mapping the distinct workflows, daily tasks, and performance metrics each profile cared about.

Information architecture was the central design challenge. The redesigned nav needed to organize 40+ feature areas into a structure that felt obvious to a first-time seller and efficient for a seller who uses the platform 8+ hours a day. Card sorting with 24 sellers across the three tiers produced the final 6-section navigation: Overview, Listings, Orders, Finance, Analytics, and Growth.

04 Solution

The ML Seller Center opens to an Overview dashboard that adapts to the seller tier. Solo Sellers see today's orders, their top 5 listings, and a single growth recommendation. Power Sellers see revenue trends, inventory risk alerts, comparative performance against category benchmarks, and a live view of pending actions across all channels.

Listing creation was redesigned as a progressive form sellers answer questions in plain language and the platform maps their answers to the required technical fields. An AI-assisted listing tool analyzes the product category and suggests optimal title, description, and attribute configurations based on top-performing listings in the same category.

Overview dashboard

Overview dashboard tier-adaptive layout for Solo and Power Sellers

Listing creation

Listing creation progressive form with AI-assisted title and attribute suggestions

05 Results

−65%

Listing creation time

+42%

Feature discovery rate

4.7

Seller satisfaction score