Marketing Campaign Optimization


An international bank developes financial products and launches marketing campaigns to potential buyers. However, past Marketing campaigns have had moderate success. Even using CRM selection techniques, it requires some manual work and final results are not satisfactory.

Marketing campaigns should be adjusted to not over saturate customers with excessive number of promotions. The approaches must select a subset of the CRM according to the probability of purchase. Therefore, a low success rate means not only losing the opportunity with potential buyers, but saturating to non-purchasers with too much information.

The company contacts Intelygenz to learn how Terminus7 Artificial Intelligence can help optimize the marketing campaigns of its financial products.

In order to measure Machine Learning success, the company carried out a first campaign with traditional methods over 291,000 customers and obtained 8,905 sales, that is, a 3% success, similar to the obtained in the past campaigns.




With demographic data and customer purchase patterns, Terminus7 uses an Attribute based approach, which establishes a Customer-Product connection based on customers’ profiles. With this approach, we look for the probability that a customer will buy a new product under campaign.

After training with several different algorithms, the team found the best Machine Learning model that best predict Campaign success.

Finally, we applied Terminus7 for a new campaign over 156,000 customers, returning 35,117 sales, which means a 23% success.

Considering the total set of 447,000 customers, from a 3% success in the first 291,000 customers, we obtained a 23% success in the following 156,000 thanks to Terminus7, giving an average of 10% success in the combined set, something unattainable with traditional CRM techniques.



Tell us your problem and we find the best AI solutions for your business