Optimizing On-Line Advertising

TL;DR

This paper develops a strategy for online advertising that maximizes revenue by dynamically optimizing ad placements over time, considering changing constraints and user interactions.

Contribution

It introduces a novel approach to optimize ad display strategies in real-time, accounting for unpredictable constraints and maximizing profit from impressions, clicks, and registrations.

Findings

Effective dynamic optimization of ad placements

Increased expected revenue through adaptive strategies

Robustness to changing constraints in online advertising

Abstract

We want to find the optimal strategy for displaying advertisements e.g. banners, videos, in given locations at given times under some realistic dynamic constraints. Our primary goal is to maximize the expected revenue in a given period of time, i.e. the total profit produced by the impressions, which depends on profit-generating events such as the impressions themselves, the ensuing clicks and registrations. Moreover we must take into consideration the possibility that the constraints could change in time in a way that cannot always be foreseen.