Customer:
The Customer is one of the top Indian organized retailers.
Expectation:
The customer was looking to build campaign response models for predicting customer response to future campaigns. These models were aimed to provide key inputs into audience selection for new campaigns.
Solution :
As part of this engagement, we developed the solution by implementing deep learning models with LSTM Architecture. Due to the large data size, the solution was implemented in distributed architecture provided by Azure Databricks.
We used customer transactional data (basket level information), historical campaigns, customer metadata to develop the solution.
Results :
The solution resulted in better targeting for the campaigns. The campaign response went up by 10% resulting in a 20% increase in incremental sales attributed to campaigns.