Case Study / Data analytics

Looking into the Future: A Simulator Predicting the Influence of Brand Drivers


Predict the drivers of Visa's brand equity to prioritize their marketing spend.

After we successfully set up a brand equity model for Visa before, they wanted to go one step further - not only identifying promising areas for action, but predicting the success of future marketing campaigns. Therefore, they needed a tool with predictive power which could be shared within different departments and across various stakeholders.

Financial Services
  • Simulator tool based on individual-level survey data
  • Proportional odds regression model for predictions

Using raw respondent data from a multi-country tracker.

To gain input for the prediction task, we used data collected in a brand tracking survey we conducted in over 20 countries and integrated it into Excel as our simulator backbone.
To ensure accurate predictions, we compared several statistical models and chose the most promising one (which was least black-box and most comprehensible). In the end, we built a user-friendly front-end within Excel that makes the simulator easy to use across Visa teams.

The simulator was jointly presented by Visa and Factworks at Predictive Analytics World in Berlin.

We provided our client with an easy-to-use simulator for cross-departmental usage.

The tool empowers marketing decision makers to simulate changes to several brand drivers’ performances and see how it affects overall brand preference. Thus, it shows areas of action where marketing budgets will most likely be well spent. 
The simulator has been presented at several conferences as a showcase of how brand driver simulations can be directly utilized for marketing practitioners.