Messi Connect
Scope
Strategy
UX/UI
Research
Testing
Branding
Type
B2C • Mobile App
Role
Lead Product Designer
Year
2025
Project Overview
Turning an ambitious AI concept into a product fans could understand, trust and use.
I led the product design from an early, difficult-to-explain and fragmented MVP to a launch-ready experience - and created foundations that could scale beyond Messi.
Registration conversion uplift from clarity-first design
Pushed notifications opted in during onboarding
Voice permissions granted from trust
Personalities can be supported by the platform
The product
What does “talking to Messi” actually mean?
Users speak naturally. The AI-driven system interprets intent, searches a library of 300–400 tagged, pre-recorded videos, then plays the most relevant Messi response.
User asks
→
Intent interpreted
→
Tagged video library search
→
Messi responds
Where we started
A strong idea built on four unvalidated assumptions.
01
Fans would understand “talk to Messi” without explanation.
02
Excitement alone would carry users through registration.
03
The voice interaction would feel self-explanatory.
04
A Messi-specific UI could later scale to other personalities.
Approach
Decision 01 · Clarity
Clarifying what fans thought they were signing up for
Hypothesis
If we showed how the interaction worked before registration, people would enter with clearer expectations and be more willing to sign up.
Field research in Camp Nou stadium · Barcelona
Key take aways
01
Excitement vs. confusion
Strong excitement and interest, but confusion about what the experience actually offered.
02
Authenticity mattered
Strong expectations for realism and personal-feeling content.
03
Guidance was necessary
The interaction model felt unfamiliar - users needed help understanding how to engage.
Concept explorations
Copy-led onboarding
REJECTED
Fast, but still asked people to imagine an unfamiliar interaction.
Full tutorial
REJECTED
Clearer, but created another long journey before registration.
Progressive demonstration
CHOSEN
Short onboarding + footage within authentication.
Decision
Design four concise moments:
Discover → Understand → Imagine → Join
then use the sign-up screen itself as a product preview.
Decision 02 · Conversion
Finding out why users dropped before reaching Messi
Hypothesis
Clear motivation plus lower perceived effort would increase the number of users reaching the first conversation.
Concept explorations
Fewer fields
EXPLORED
Reduced friction but kept the least familiar route dominant.
Progressive steps
REJECTED
Each screen looked lighter, but interaction made the journey feel longer.
Compact fields + Familiar patterns
CHOSEN
Apple, Google or email; one dominant action per state.
VALIDATION
Compared structures, layout and messaging in prototypes and tested them in Maze; then monitored released funnel performance.
OUTCOME · +26%
Registration conversion improved by combining familiar pattern registration with reduced effort.
Decision 03 · Interaction
Making an unfamiliar conversation feel familiar
Hypothesis
Video-call conventions plus explicit system states would let users speak with less instruction.
Concept explorations
Chat first
EXPLORED
Immediately understandable, but weakened Messi’s presence.
Content first
REJECTED
Controlled and cinematic, but too far from the promise of conversation.
Conversation first
CHOSEN
Full-screen video with explicit listening and response cues.
Decision 04 · Trust
Turning permissions into a value exchange
Hypothesis
If people understood the benefit before the operating-system prompt, they would be more likely to grant access.
BEFORE
Immediate native prompt
Appeared before users had a reason to understand why they would need it.
AFTER
Contextual pre-permission
“Speak naturally to interact with Messi.” Native prompt follows only after the value is clear.
67% NOTIFICATIONS
Opt-in when linked to new stories and limited access.
94% MICROPHONE
Permission acceptance after contextual explanation.
Decision 06 · Scalable
Designing Messi Connect without designing only for Messi
Separate the stable product layer from the changeable personality layer by the Design system themes to make a white-label product.
Product layer
Navigation
Registration and permissions
Conversation flow
Content structures
Component behaviour
Personality layer
Colour and typography
Imagery and video
Branded assets
Theme tokens
Bringing it together
I tackled the design goals across three areas:
01
Redesigning the core user flows
02
Creating interaction structures that help users understand a novel experience
03
Building a comprehensive design system
Design Goal: Conversion
A strong Store presence
Bold visuals and clear positioning made Messi Connect unmistakable on the App Store.
Across 25 weeks (27K+ impressions):
11% average CTR (18.4% peak)
4.2% conversion to download (8.3% peak)
Refined through A/B testing and keyword localization (EN/ES).
Design Goal: Conversion
Gamifying the experience
Messi's content wasn't recorded yet at launch. To keep users engaged, I designed a progression system that rewarded exploration: match highlights creating natural re-entry points, an interactive globe mapping iconic moments to career milestones, and a set of collectable stickers that could be unlocked through replays.
This built habit and anticipation.
Building a cohesive brand across touchpoints
Design Goal: Clarity & Conversion
I expanded the product experience beyond the app, designing the website and social presence to create a consistent, recognisable identity across platforms.
Aligning visual language, messaging and tone helped reinforce the concept, build trust early, and drive engagement from the first touchpoint.
Help users understand a novel product. Guide them through sign-up, set expectations and explain the interaction model.
Impact
+26%
Registration conversion uplift from clarity-first design
+67%
Pushed notifications opted in during onboarding
+4.2%
Conversion to download rate
+94%
Voice permissions granted from trust
Infinite
Personalities can be supported by the platform
Learnings
01
Reducing friction drives engagement
02
Trust is part of the experience
03
Systems thinking scales products
This project was full of firsts for me. It was my first time designing an AI-powered product and my first time building a design system that could support multiple brands. It pushed me to think beyond the immediate MVP and design foundations that could evolve as the product grew.
I also learned that successful products aren't defined by how many features they include, but by how quickly users experience value.
Designing the onboarding journey reinforced the idea that reducing friction has often a greater impact than adding functionality. Every decision; from simplifying registration to explaining permissions, was an opportunity for us to build trust and help users reach the core experience faster.
