A new approach to data-driven marketing implements media mix modeling and unlocks exponential growth at Meta. Learn about the success story.

Context: a simpler, closer and 100% digital bank

Galicia is one of the leading banks in the Argentine financial system and one of the most prestigious banking institutions in Argentina. With a user-centric approach, it offers simple, reliable processes designed for its customers' everyday lives.

In the competitive world of finance, every advertising investment counts. Therefore, Galicia partnered with Meta and Bunker DB to optimize its marketing investment and boost the performance of its campaigns. Its main objective was to increase new customer acquisition through a data-driven approach.

Challenges: measure better to invest better

Galicia sought to take a holistic and precise look at how its different marketing channels interacted and identify the incremental impact of each. This way, they could efficiently scale campaigns and significantly increase new customer acquisition.

However, Galicia's marketing team faced several challenges:

Dispersed data

Their information was fragmented across online, offline, and CRM channels, with different formats and granularity levels.

Complex metrics

The financial sector goes beyond clicks or conversions. It is powered by models that capture customer lifetime value, lead-to-customer conversion rates, and the average value of each customer.

External factors

The financial decisions of Argentine consumers can be influenced by macroeconomic factors, currency volatility, and changing interest rates, adding complexity to any analysis and optimization.

Solutions: towards a better attributed holistic analysis

Galicia turned to Meta and its measurement partner Bunker DB to reduce bias, improve its attribution systems, and uncover the true contribution of each channel. After identifying multiple challenges, the following actions were taken:

Data Normalization and Unification

To unify all the different metrics, an ETL (Extract, Transform, and Load) process was implemented, consolidating and standardizing data from online, offline, and CRM sources. In addition, advanced interpolation and aggregation techniques were applied to harmonize different granularities and ensure information consistency.

Share of Voice Analysis

A variable was constructed to evaluate the brand's presence relative to competitors in digital and traditional media, using social listening to analyze brand perception and online conversations.

Marketing Mix Modeling (MMM)

Robyn, Meta's open-source solution, was implemented to quantify the relationship between marketing investment and customer acquisition. The model incorporated exogenous variables such as interest rates, economic indicators, and seasonal factors to improve the accuracy of its predictions.

Results: a more realistic reading of the impact of each channel

The results obtained with Robyn were overwhelming. During Q1, Meta-driven account openings increased by 50%, while reducing the average CPA by 10%, demonstrating the impact of the short-term improvements implemented in the media mix.

And most importantly, a more realistic reading of the impact of each channel, especially digital channels, which proved to have a greater effect than initially estimated.

Next steps: a new mindset for bank marketing

The implementation of the model changed the way Galicia makes advertising decisions. Today, marketing is managed with a focus on real impact and a holistic approach that involves the entire business.

After implementing the model, Galicia's next steps will be to conduct incrementality studies such as Conversion Lift that will help measure the incrementality of each channel and campaign, calibrate the model, and seek new efficiency hypotheses.