Identifying customers with highest ROI on ads

With almost 75% of people carrying a smartphone in their pockets, online marketing has become a heavyweight customer acquisition channel, especially when considering deepsense.ai’s client, a leading CEE fashion retailer.

The challenge

Harnessing the wealth of data available today leads to significant cost and performance optimizations. However, when not targeted properly, online advertising can lead to significant losses. This is true particularly when ads are shown to two types of customers: those who are completely uninterested in buying the product being featured and those that will buy it without being exposed to the ads in the first place.

Experiments have shown that displaying ads to customers most eager to buy a product is a waste of money as they will buy regardless of seeing the ads. There is also a group that will not buy the product regardless of the number of ads seen. Thus, only the group in the middle should be exposed to ads to make the campaign more profitable – displaying the ads to the customers extremely willing to buy or completely uninterested is practically wasting the resources.

The solution

deepsense.ai delivered a standalone, cloud-based solution that scans through real-time data from popular e-commerce systems, including Magento platform and Google Analytics.

The solution uses machine learning algorithms to identify customers with the highest ROI on internet marketing. The retailer increased its profits from advertising while reducing wasted resources.

The effect

The predictions our model delivers are accurate in 80% of cases and the most valuable group of clients witnessed a 2,5x higher uplift from advertising.