See all references

Purchasing propensity in the automotive sector.

Use case


In this scenario, the client is an automotive brand with a large number of followers, making it challenging to distinguish between those who are merely fans and those who are genuinely going to purchase a vehicle.

Within its customer-centric philosophy, it pays great attention to anyone interested in its vehicles and the brand., but there is a need to analyze the propensity of its current customers to acquire new vehicles, both for those customers who already own a vehicle from the brand and for potential leads.


This purchasing propensity initiative enables the identification of those customers and leads who have a higher probability of buying each specific model of the brand.

On one hand, a discovery of variables will be conducted, characterized over time and by location, that have a greater predictive capacity based on the historical purchase data, lifestyle, behavioral segments, proximity of products, or socio-demographic information concerning the studied vehicle.

Customer clustering based on similarity and study objectives will be performed by analyzing their interactions with the entity over a historical period of 10 years.


Next, there will be a training and validation process for algorithmic models to obtain the probability of the customer purchasing the target vehicle. This involves generating a probability scoring for each customer based on the target model, finishes, and the acquisition model.
The initiative was approached in two phases based on the collected information: one focused on existing customers of the brand, and another that incorporates information from leads added to the CRM.


  • Scoring of entity’s customers on acquisition propensity for 19 target models/finishes is performed based on over 350 variables for each customer.
  • Improvement in results is achieved over the traditional model used by the brand for calculating propensity to contract in the majority of the set objectives.