Media agency

Garage Model: An automated dashboard, delivered by a team that didn't code.

The way of working we created at a media agency to stand up a dashboard for media and social campaigns: fifteen days to understand the client and forty-five to deliver.

• Sector: Media agency
• Timeframe: 15 + 45 days
• Team: Agency staff, no coding

Tools and methods: dataviz, big data, SQL, Python, PMI and agile methods.

Each platform with its own numbers, put together by hand in a report. A single data source for the client and the agency.

A common request that nobody had solved.

The client asked for an automated dashboard for its media and social campaigns, with insights for the client and for the media teams. In agencies, data automation is a constant request, but people only talk about the need.

A data process is much more than building databases, in the cloud, in SQL or in spreadsheets, especially when the data is sensitive.

Fifteen days to understand, forty-five to deliver.

We created the Garage Model from PMI models and agile and rapid prototyping methods, to deliver scalable concepts quickly and repeat MVPs in data, insights and automation projects.

• Assessment: Receive the client's brief and documentation.
• Mapping: Develop the measurement plan and set up the tools.
• Build: Design the report for the people who will use it.
• Operation: Automate and iterate, with collection, analysis, interpretation and a weekly report.

After delivery, the cycle repeats every week.

The last phase is to automate and iterate: collect the data, analyse it, interpret it and close with the weekly report. Each round improves the next.

• Collection
• Analysis
• Interpretation
• Weekly report

After the report, the cycle goes back to collection.

In marketing, data is not just automation.

It is a culture that changes with people in mind and gives them autonomy. The model rests on four pillars.

• Relevant: A process architecture that controls requests and develops the team.
• Secure: A secure data ecosystem.
• Targeted: Focused dashboards, designed for the people who will use them.
• Effective: Faster insights for the client and for the agency.

What we chose to do.

• Process before tools.: A data process is more than databases, cloud, SQL or spreadsheets. We started with the measurement plan and only then set up the tools.
• The team builds, it doesn't just receive.: The automation was delivered by a team that knew neither programming languages nor these methods. It left the project able to run the next ones on its own.
• MVP first, scale later.: Agile methods and rapid prototyping to deliver a working concept quickly and repeat it in other data, insight and automation projects.
• Designed for the people who use it.: The dashboard was designed around the people who would read it, the client and the media teams, not around the data that was available.

In 60 days, the automation and a self-sufficient team.

The agency, in other countries too, recognised the proposal as a new business model for large clients.

The result was a single data source, with more agile processes and data available to the whole team. Relationships with partners and clients became closer, customer experiences more effective and operations more efficient.

The project led to others.

It also brought the team closer. Focusing on people's development, with clear pillars, is what made everyone involved grow.

From Brazil to IPG Global It was the project we took global at IPG. The person who presented it, now part of our team, was the only Brazilian at IPG to do so.