Senior Product Designer
The only PD across the entire customer row at Daki
The team
Product lead Brazil, Engineers, Jokr and Daki CTOs and VPs
Responsibility
End to end product design and rollout support for the Brazil market
The Problem
We needed to merge the existing Daki app with JOKR into one global experience. The goals were to maintain or improve conversion and retention during a forced migration, avoid churn from Brazilian users, and align to global systems and constraints.
Key notes
Merge two live grocery delivery apps into a single global product without hurting core business metrics in Brazil.
Operated as the only Product Designer at times, partnering with a newly forming leadership and an engineering squad from an external consultancy.
Built a migration plan that preserved the Daki user base, aligned to the new global strategy, and review the design system for scale.
Result: conversion held parity, the entire Daki user base migrated to v1 of the new app in 16 weeks, AOV increased by 3%, and huge add-to-cart clicks increased
Company and Product
Daki is a mobile first ultra fast delivery platform in Brazil. In 2022 Daki joined JOKR and both apps were consolidated into a single platform. While the core business overlapped, the apps differed in architecture, user journeys, and market specific features.
Constraints and Challenges
Different backends and design systems between apps
Brazil specific payments and address formats (Pix, CEP, building number complements)
Checkout flow focused on coupon reassignment based on a given MOV or strike-through rule
User perception about product discovery flow considering the change in navigation in the categories
Tight timeline with hard business milestones
Live ops pressure: orders, promotions, and substitutions running daily
App architecture
Mapped the full navigation structure, defined context segmentation, and documented how each app’s business metrics relate to one another for Daki and Jokr.
Discovery and audit
Compared end to end journeys from both apps using task time studies and funnel data, mapped gaps for Brazil: payments, address formats, fees, taxes and bundles.
Design system harmonization
Tokenized colors, spacing, and typography to fit two different brands (Jokr and Daki). Component parity checklist across both apps and creation of new components.
Validation strategy
Maze researches, moderated perception test with Daki's heavy users (High Value Price Seeker and High Value User) and AB Tests.
Outcomes
Checkout conversion: parity (no change)
Meaningful improvement in add-to-cart behavior
Increased Average Order Value (AOV) by 3%
100% of the Daki user base migrated to v1 of the new app in 16 weeks
What I Would Do Differently
It was no stable product trio, decision makers shifted week to week while an external consultancy owned much of delivery. We didn’t have time to conduct research, but we did it anyway with relevance and quality. However, in the process I could have simplified some decisions related to the research.
