Emerge's cross-sell models help sell more products.
This case study is for one of South Africa’s big banks. Emerge built three machine learning models using transactional account data and data from three other products to identify clients who did not yet have those products but were prime candidates to buy up. The results for the 3 models are as follows:
- The results were most exciting for the cross-sell to insurance process- from approximately 3% of the transactional account base, it was predicted that 70% of clients would buy an insurance product
- For investment products the equivalent results showed that for about 15% of the current account base, 60% of clients would buy up to investment products
- Finally, for the cross-sell campaign to the bank’s loyalty product, showed that about 11% of the current account was needed to find 40% of clients that would buy up.
These models are yet to be implemented. The expected value they are expected to create is in excess of R5m per month.