Beverages · Brazil

Consumer data journey

A proposal for a large beverage company in Brazil: build a data structure to identify consumers and activate them strategically, with a customer data platform and a 360 view for marketing.

Sector: Beverages · Brazil
Format: Proposal
Method: Garage Model · 2-week sprints
Tools: Tableau · Python · big data

The request

The brand teams needed data they could use day to day. That means combining three types of data and linking them to the journey stage each consumer is at: discovery, consideration or conversion.

First-party data: brand affinity and product interest, such as juices, beer and energy drinks.
Demographic data: location, age, gender and life stage.
Audience interests.

Three principles

Actionable data for the brand teams, based on three principles.

Customer-centric: Unify audience data.
Efficient: Automate what repeats.
Agile: Activate quickly.

Customer data platform

The first step is to take control of the technology and keep the data together. A customer data platform brings all sources into a single view and feeds dashboards, segmentation and predictive models.

Personal and demographic data, website behaviour, engagement, transactions and devices.
First-party and third-party studies.
Business, marketing, sales and digital experience dashboards, with segmentation and predictive models.

Diagram of a customer data platform that combines several sources and feeds dashboards, segmentation and models — Unified view, from data sources to dashboards and models · original presentation

360 view

The second step is to find the signal in the noise. The proposal set out a bespoke platform, with fresh data for marketing, in four phases: dashboards, machine learning, publishing and insights.

Dashboards: explore the databases and build Tableau dashboards with a key user.
Machine learning: choose the KPIs that justify custom algorithms.
Digital Score: a ranking of users by engagement.
Recommendation system: segments by affinity with content and products.

Four phases: dashboards, machine learning, publishing and insights — The platform’s four phases · original presentation

Discovery

Before building, eight weeks of discovery: kick-off, co-creation session, audit and research, through to defining and prioritising the MVPs and agreeing the implementation plan with the client.

3 weeks to set up the team and access.
Audit and research in the first weeks.
Definition and prioritisation of the MVPs, implementation plan and client approval.

Eight-week discovery schedule — Discovery plan, week by week · original presentation

Sprints

Implementation runs in two-week sprints, with a strategy squad and an engineering squad. Each sprint starts from user stories, validates the databases and ends with a review and an MVP release.

Sprint cycle: backlog, planning, daily stand-ups, review and retrospective — Sprint work cycle · original presentation

Benefits

A single view of the data brings gains for the whole organisation, not just marketing.

A single source of truth.
Data accessible to the whole team.
Agile, simple processes.
Closer relationships with partners and suppliers.
More effective customer experiences.
Operational efficiency.