Bidding Machine: Learning to Bid for Directly Optimizing Profits in Display Advertising

TL;DR

This paper introduces Bidding Machine, a unified learning framework for real-time bidding in display advertising that jointly optimizes utility estimation, market value forecasting, and bid decision-making, leading to improved campaign performance.

Contribution

It presents a comprehensive joint optimization framework for RTB bidding that integrates three interconnected components, unlike previous methods treating them separately.

Findings

Joint optimization significantly improves campaign effectiveness.

The framework can be trained with offline batch or online sequential schemes.

Extensive offline and online tests confirm high effectiveness.

Abstract

Real-time bidding (RTB) based display advertising has become one of the key technological advances in computational advertising. RTB enables advertisers to buy individual ad impressions via an auction in real-time and facilitates the evaluation and the bidding of individual impressions across multiple advertisers. In RTB, the advertisers face three main challenges when optimizing their bidding strategies, namely (i) estimating the utility (e.g., conversions, clicks) of the ad impression, (ii) forecasting the market value (thus the cost) of the given ad impression, and (iii) deciding the optimal bid for the given auction based on the first two. Previous solutions assume the first two are solved before addressing the bid optimization problem. However, these challenges are strongly correlated and dealing with any individual problem independently may not be globally optimal. In this paper, we propose Bidding Machine, a comprehensive learning to bid framework, which consists of three optimizers dealing with each challenge above, and as a whole, jointly optimizes these three parts. We show that such a joint optimization would largely increase the campaign effectiveness and the profit. From the learning perspective, we show that the bidding machine can be updated smoothly with both offline periodical batch or online sequential training schemes. Our extensive offline empirical study and online A/B testing verify the high effectiveness of the proposed bidding machine.