Audience
Strategy

SKY had the data and nobody had translated it into decisions. I read the signals and wrote the taxonomy: three audiences, three homepages.

Audience Strategy Experience Architecture Data Design Behavioral Analysis Personalization
Client SKY Brasil
My Role Senior UX · Audience Strategy and Experience Architecture
Role scope I interpreted the behavioral data, defined the taxonomy, and translated signals into strategy that data and engineering implemented via the DMP.
Year Before 2023
Outcome Three clusters mapped, personalized flows deployed, and first-party data strategy documented in the DMP.
3 Audience segments
1st First-party data interpreted
DMP Data strategy
Personalization Personalized flows
Audience StrategyData DesignBehavioral AnalysisExperience ArchitectureSKY BrasilPersonalization Audience StrategyData DesignBehavioral AnalysisExperience ArchitectureSKY BrasilPersonalization
SKY Brasil
Section 01
Context

Data without context is just noise.

SKY Brasil had the infrastructure. What was missing was context: nobody had translated the behavioral data into experience decisions.

SKY already ran a DMP collecting first-party signals. The infrastructure existed, interpretation was missing.

◎
The Design Opportunity

The DMP let us personalize the site itself: each profile seeing its own navigation and hierarchy. The challenge was defining what each variation contained.

→
The Translation Layer

Behavioral data only drives experience when someone writes the rules. My contribution was the taxonomy: which segments matter and what UX response each requires.

∑
Strategic Progression

Behavioral data → Audience segmentation → Experience strategy → Personalized flows. Each layer depended on the clarity of the one before.

Behavioral clustering framework
Behavioral clustering framework: mapping data signals to audience structure
Section 02
Research

Reading the signal
as intent.

Alignment first. Then a deep read of signals across four data layers.

I started by aligning strategy, design, and data science on the subscriber journey. The question wasn't what they clicked, it was why.

I worked through funnel drop-off, intent signals, and flow between content areas. Someone arriving via sports wasn't the same customer as someone arriving via price, but the site treated them as if they were.

"Map affinity by crossing aspirational signal with attitudinal data: what people wanted versus how they actually decided."
01
Purchase Funnel Analysis

I mapped drop-off points and diagnosed the cause: content, UX, or wrong audience.

02
Organic Traffic Intelligence

Analysis of the search terms driving organic traffic. Identification of intent signals that revealed radically different motivations landing on the same homepage.

03
Behavioral Flow Mapping

I tracked paths between content areas, isolating what distinguished a converter from a visitor.

04
Attitudinal Layer

Cross-referencing behavioral data with content engagement patterns, sports followed, and entertainment consumed, to build the reasoning behind the behavioral signatures.

Behavioral data interpretation diagram
Behavioral data interpretation: mapping signals to audience intent
Behavioral flow analysis and analytics data
Behavioral flow analysis: analytics data driving audience segmentation
Section 03
Audience Model

Three audiences. Three experiences.

The data resolved into three distinct behavioral clusters, each with a different decision model and a specific UX requirement.

The data resolved into three behavioral clusters, each with its own decision style and UX requirement. Not marketing personas, interaction models, each mapped to a different homepage and conversion path.

01
Segment 01
Opportunistic

High-intent, price-sensitive. Arrives via promotion or comparison. Needs immediate clarity on the offer. Decides fast.

Conversion: featured offer → plan comparison → subscribe
02
Segment 02
Value-Driven

Deliberate and analytical. Compares plans, reads the FAQ, calculates cost per channel. Needs comprehensive information architecture.

Conversion: plan configurator → FAQ → contact or subscribe
03
Segment 03
Content Lover

Aspirational, entertainment-first. Arrives via film, not product search. Needs a content-led experience.

Conversion: content discovery → what's included → subscribe
Audience segmentation model
Segmentation model: behavioral signals mapped to clusters and experience variations
Section 04
Experience Strategy

From cluster to creative strategy.

With the three segments defined, I built decision frameworks for each, converting behavioral logic into DMP rules and interface variations.

I wrote a creative strategy per cluster: tone, content hierarchy, visual structure, and the flow the DMP would serve each profile. Not wireframes, decision frameworks that technology turned into rules and design executed as interface.

Design Principle 01
Segment at entry, not at checkout. Personalization that only fires at conversion is too late. I designed the DMP to read behavior in the first two page views and adapt navigation from there.
Design Principle 02
The homepage is not one page. SKY's was three, served to three audiences. The work was defining what each version contained and which DMP rule triggered it.
Design Principle 03
Data must be classified before it can be creative. The most important document I produced was the matrix translating behavioral signal into audience.
Personalization strategy showing audience segments triggering different navigation experiences
Personalization flows: how each segment triggers a distinct navigation experience
Section 05
Operational Impact

Strategic outcome with measurable reach.

The audience strategy produced operational results that extended beyond the UX scope, reaching the data, marketing, and technology teams.

The taxonomy and experience rules became shared infrastructure across data, marketing, and technology.

Homepage personalized by behavior: the DMP began serving three distinct navigation architectures.
Less mismatch between intent and navigation: the architecture started following each cluster's real decision model.
Shared taxonomy: the matrix became the operational reference for marketing, data, and creative.
Documented UX rules: which signal triggers which variation, making personalization reproducible.
A model established for data-driven UX. The methodology became a replicable model for subsequent personalization initiatives.
Retrospective

What this project taught me
about design and data.

This project was about translating complexity into clarity, and clarity into decisions.

The instinct in data-heavy projects is to defer to the analyst. The hard part is constructing the questions that make data meaningful.

Lesson 01 · Data Literacy
UX strategy requires data literacy, not data dependency
I read analytics platforms, interpreted flows, and built taxonomy from first-party signals. The capability wasn't data science, it was disciplined translation.
Lesson 02 · Organizational Design
Personalization is an organizational capability, not a feature
The strategy only worked because data, creative, and technology shared the same taxonomy. Personalization at scale is organizational capability, not technical.
Lesson 03 · Strategic Visibility
The most strategic design work is often invisible
My highest-impact deliverables were entirely conceptual. That shaped how I demonstrate strategic UX value: by outcome, not output.
Next Project
SAMSUNG · Infinite Island
Experience Strategy Digital
→
 Back to top