Effectiveness of Automatic Retargeting in Programmatic Advertising

Abstract

Advertising platforms such as Facebook and Google provide the advertisers with tools allowing them to benefit from advanced targeting, automated ad delivery mechanisms, and behavioral algorithms. These algorithms are based on machine learning and can predict whether a user will click on a particular ad, allowing for optimization of their placement. Online marketing is transitioning to automatized lead nurturing strategies that follow the customer along his journey with a series of personalized and automated messages. A significant part of this process is represented by dynamic retargeting. The method of dynamic retargeting is used in advertising to address users that already interacted with the brand online and offer them personalized content to motivate them to proceed in the next stage of the customer purchase journey. While automation is generally associated with the increased effectiveness of the processes it is applied to, the human aspect of marketing creates a different context with unexpected responses. A growing corpus of studies questions the effectiveness of targeted online advertisements in relation to the increasing number of adblocker users and decreasing click-through rates. However, these studies usually do not distinguish the type of advertisements or the purchase stage. Dynamic retargeting represents an advanced targeting strategy that uses a combination of cookie- and pixel-based user tracking and personalized offers using e-shop generated product feeds. This article tests the effectiveness of 216 sets of dynamic retargeting ads by comparing them with 216 non-dynamic, manually created product ads both in Facebook Ads Manager and in Google Ads.

Keywords: Dynamic retargeting, Facebook ads, Google ads, hyper-targeting, marketing automation, online advertising, personalized advertising, social media

JEL Classification: M30,M31,M37