Marketing Campaign

A successful marketing campaign for advertising of various kinds of products involves the provision of relevant advertising to the user of the Internet resource on which he is located, which in turn is associated with the exact definition of the category of client and his preferences. A topic model of user behavior profiles within an Internet resource is built on the basis of transactional data of interaction with the site from cookies.

Input Data

  • List of URL-pages, which the user has visited
  • Description URL-pages
  • Accompanying user information
  • User ad results

Output Data

  • Model the behavior of Internet users of a resource
  • Prediction model of the response to an advertising campaign
  • Interpretation of customer base

Main Results

A pilot project was carried out, integrated into the customer’s business process:

  • Identified themes 80% of the user visited pages.
  • 70% are interpretable
  • Prediction of socio-demographic parameters of the audience
  • The interests of 90% of visitors by topics of visited pages are highlighted
  • 80% of interests are interpretable
  • Increased advertising campaign conversion by 10% with A / B testing
  • Changes made to 15% of finished advertising campaigns

The final quality of the selection of customer behavior profiles for increasing the conversion of an advertising campaign:

Quality criterion Quality
The proportion of interpreted topic URL-pages 70 %
The proportion of interpretable behavior profiles 80 %
F1-score forecasting social-dem. (gender) 68 %
MAE forecasting social-dem. (age) 8,01

Quality built soft clustering system

Clustering quality criterion Quality
Accuracy 83 %