Customer profiling from web browsing data

Use case

Regardless of the product, one thing is common for every marketing campaign – the more tailored the campaign is to the profile of the customer, the higher the chance of successful conversion. In this project we used machine learning for customer profiling based on their web traffic.

O2 Slovakia is one of the major telco operators in the country. Its data warehouses contain terabytes of data and our team was called in to extract the value contained in it.

Solution

We constructed a machine learning solution that processes semi-structured text data obtained from mobile web browsing of their users. Based on this data, the tagging engine assigns some of the content categories to each of the websites a user visited, therefore providing an in depth insight to the customer’s profile.

Result

Thanks to a high level of automation and skill of our team, we have created a prototype of the solution in less than 20 days and handed it to the customer’s in house data science team. The customer obtained a high quality solution at a symbolic cost.