Redesigning the product catalog for faster, smarter sales
A complete redesign of the product catalog feature on Mercado Pago's Point Smart device: making it faster, more flexible, and accessible to small and medium-sized businesses across Brazil.
The challenge at a glance
Point Smart is Mercado Pago's flagship payment device for physical stores — a touchscreen POS that lets merchants accept card payments and manage sales. The device includes a product catalog feature that allows sellers to build a cart by tapping items, rather than manually entering amounts.
Despite its potential, the feature was barely used. Only around 1 in 10 active users ever touched it. The goal of this project was to understand why, and redesign the experience from the ground up.
What users actually needed
I led user interviews and usability testing sessions with merchants across different business sizes and contexts. Four core themes emerged that shaped every design decision that followed.
Sellers needed to mix catalog items with custom freeform amounts in a single transaction — the existing flow forced a binary choice between the two.
Merchants with large inventories faced a steep entry barrier. Setting up hundreds of items manually was so time-consuming, most never started.
Cashier staff needed quick visual cues to find items fast. Text-only lists created friction at the moment of sale, especially with multiple people using the same device.
The primary use case happened at the counter, often with customers waiting in line. Every extra tap or scroll was felt acutely in real conditions.
Aligning user needs with business objectives
The project had three interlocking business goals that we balanced against the research findings throughout the design process.
Engagement
Increase the perceived value of the catalog feature among existing users — making it something merchants genuinely relied on, not just a tab they ignored.
Retention
Increase switching costs by making the catalog deeply integrated into how merchants run their day-to-day operations — creating stickiness through utility.
Move-up market
Build a feature powerful enough to serve the SMB segment — businesses with larger, more complex inventories who had been unable to use the feature at all.
What made this project genuinely hard
Beyond the design brief, the team navigated three layers of real-world constraints that shaped our decisions throughout.
User range
We had to design for a spectrum from a street vendor with five items to a restaurant with hundreds of SKUs — a single interface had to serve both without compromise.
Retro-compatibility
The feature was already live in production. Every change had to be backward-compatible with existing user item lists — we couldn't break what was already working for the 10% who used it.
Roadmap alignment
We were designing not just the current state but a foundation for upcoming features. Every decision had to accommodate a broader product roadmap without over-engineering the present.
Designing in the real world
The process followed a listen → analyze → design → test loop — though in practice it looked less like a clean diagram and more like overlapping, iterative cycles. I led the design effort end to end, working closely with product, engineering, and business stakeholders at each stage.
Understand before moving
Started with a deep dive into existing analytics, product goals, and technical constraints before touching any design tools. User interviews followed, conducted with merchants across different business sizes in Brazil.
Progressive fidelity
Design proposals evolved in detail alongside the project — starting with rough sketches and concepts, gradually increasing fidelity in Figma as decisions were confirmed and scope was locked. Avoided premature high-fidelity work that would need to be thrown away.
Facilitating alignment
Every design review was structured to present the problem, the options considered, and the reasoning behind the chosen direction — not just the final screens. This built cross-functional trust and kept decisions moving forward.
Usability testing
Prototypes were tested with real merchants using Figma interactive prototypes. Sessions focused on speed and accuracy at the point of sale — the high-pressure moment the feature needed to serve.
Launch and follow-up
Treated launch as the beginning, not the end. Monitored adoption metrics and NPS post-release, using early signals to inform the complementary features roadmap.
Key design decisions
Mixed-sale flow
One of the most impactful decisions was redesigning the transaction flow to natively support mixed carts — combining catalog items and custom amounts in a single checkout. Previously these were two separate, incompatible modes. The new design unified them through a shared cart model accessible from both the value entry screen and the catalog tab.
Color logic for quick item access
Research showed that cashier staff needed to identify items fast — especially in busy, counter-top environments. We introduced a color assignment system for both categories and individual items. Merchants could assign colors during setup; at point of sale, the visual language reduced scan time and cognitive load significantly during usability testing.
Items without photos display color-coded tiles with initials. Items with photos display the image, maintaining visual consistency across mixed catalogs.
Countertop interaction model
A subtle but important constraint: the device typically rests on a countertop, used at a slight downward angle. This informed touch target sizing, the placement of primary actions at the bottom of the screen (closer to the user's natural reach), and the decision to make the cart button persistent and always visible during browsing.
Designing for scale
The catalog grid needed to work for 5 items and 500. Category filters with horizontal scrolling allowed merchants to navigate large inventories without overwhelming the primary view. Search was added as a fallback for power users with extensive catalogs.
Outcomes after launch
The redesign shipped to users in Brazil and was monitored closely in the weeks following launch. The results validated the research-led approach and opened the door to the SMB roadmap.
What came next
The redesign was the foundation for a broader set of features aimed at making the catalog more powerful for larger businesses. Three initiatives were scoped as direct next steps.