Platform Initiative
Communication Infrastructure Platform
Building the customer communication platform of one of Brazil's largest banks.
Itaú Unibanco · Notifications & Preferences · 2022–2024 · Channel PM
The Product
Itaú's central customer communication platform. Concentrates every conversation (notifications, approvals, messages) into a single, simple and humanized surface.
The Scope
Notification and preferences channel for every retail segment: Retail, Uniclass, Personnalité — and later iti and Cards via the One Itaú strategy.
My Role
PM for the Notifications & Preferences channel. Owned everything from initial discovery to strategic deliveries such as IDL v2/v3, Salesforce SDK and GA4.
Pain points identified in the product
- Decentralized communication journeys across the app
- Notifications surfaced in unfriendly patterns
- Approvals hidden in hard-to-find areas
- Complex and fragmented preference management
- Multiple bank apps communicating in inconsistent ways
Business challenges
Product serving every retail segment of the bank simultaneously — Retail, Uniclass, Personnalité, iti, Cards.
One Itaú — unifying the experience of multiple apps into a single superapp, with the Hub as the central channel.
IDL — evolving the product to the bank's new visual standard without breaking the existing experience.
Integrating the Salesforce SDK for personalized campaigns while preserving channel performance and availability.
Execution approach
Discovery
Mapped real customer pain: decentralized journeys, unfriendly notifications and complex preferences. Defined squad purpose and OKRs: +70% click-through and 75% PF opt-in.
One Itaú
Absorbed iti and Cards notifications into the hub. Coordinated multi-channel integration while preserving experience consistency across all retail segments.
IDL v2 → v3
Created a unified hub (v2) with notifications, messages and pending actions. After Personnalité beta, separated pending+notifications from messages (v3) — a decision driven by real user behavior.
Integrations
Implemented the Salesforce SDK (Android and iOS) for personalized campaigns. Led the GA4 migration and evolved the Preferences journey with unified storage under a single key (FE9).
The Most Important Decision
IDL v2 → IDL v3
Initial hypothesis (v2)
A unified hub with notifications, messages and pending actions in a single experience. The logic was: everything in one place = more simplicity.
Problem surfaced with Personnalité beta testers
Messages and notifications had completely different intents. Mixing them created confusion — not clarity.
Data-informed decision (v3)
Separated pending actions and notifications into a single tab and routed messages to a dedicated icon in the app header.
Outcome
- ✓App navigability improved.
- ✓Each communication type with its appropriate space.
- ✓More intuitive experience validated with real users.
- ✓Shipped as a usability improvement to the Lab.
"A good hypothesis is one you're willing to refute. The best decision I made was to change my mind based on what users actually did."
Operational systems created
Discovery & vision
Squad purpose, vision and OKRs defined from zero. Pain and opportunity mapping.
Notification hub
Centralized all customer communications into a single channel with time-based grouping.
IDL v2 → v3
Design system evolution: unified hub and intelligent split between notifications and messages.
Salesforce SDK
Full iOS and Android integration for segment-personalized campaigns.
Preferences
New preferences journey with unified storage (FE9) — customers control their own channels.
GA4 & analytics
Tagging migration aligning the bank with Google guidelines and enabling advanced analytics.
Outcomes & learnings
Core realization
Platform product is fundamentally different from business product — your customer is internal, but the impact is on the end user. I learned to balance the needs of multiple consuming squads against the channel's long-term vision, and to make product-architecture decisions based on real user behavior, not assumption.
Key operating principles
Platform products have two customers
Internal (squads) and end (app users) — balancing both is the work.
A good hypothesis is refutable
IDL v3 was born from changing my mind based on data, not opinion.
Cross-product strategy
One Itaú requires active coordination — absorbing isn't enough; integrating with quality is.
Experience architecture is product
Separating notifications from messages was as important as any feature.
OKRs orient prioritization
With clear OKRs from day one, the team knows what to move and why.