Dynamic bidding strategies in search-based advertising

Abstract

Search-based advertising allows the advertisers to run special campaigns targeted to different groups of potential consumers at low costs. Google, Yahoo and Microsoft advertising programs allow the advertisers to bid for an ad position on the result page of a user's query when the user searches for a keyword that the advertiser relates to its products or services. The expected revenue generated by the ad depends on the ad position, and the ad positions of the advertisers are concurrently determined after an instantaneous auction based on the bids of the advertisers. The advertisers are charged only when their ads are clicked by the users. To avoid excessive ad expenditures due to sudden surges in the keyword-search activities, each advertiser reserves a fixed finite daily budget, and the ads are not shown in the remainder of the day when the budget is depleted. Arrival times of keyword-search instances, ad positions, ad selections, and sales generated by the ads are random. Therefore, an advertiser faces a dynamic stochastic total net revenue optimization problem subject to a strict budget constraint. Here we formulate and solve this problem using dynamic programming. We show that there is always an optimal dynamic bidding policy. We describe an iterative numerical approximation algorithm that uniformly converges to the optimal solution at an exponential rate of the number of iterations. We illustrate the algorithm on numerical examples. Because dynamic programing calculations of the optimal bidding policies are computationally demanding, we also propose both static and dynamic alternative bidding policies. We numerically compare the performances of optimal and alternative bidding policies by systematically changing each input parameter. The relative percentage total net revenue losses of the alternative bidding policies increases with the budget loading, but were never more than 3.5 % of maximum expected total net revenue. The best alternative to the optimal bidding policy turned out to be a static greedy bidding policy. Finally, statistical estimation of the model parameters is visited. © 2013 Springer Science+Business Media New York

Multi-objective cluster based bidding algorithm for E-commerce search engine marketing system

“…Game-theoretic structures of SEM have been studied by Borgers et al ( 2013 ) and Aggarwal et al ( 2009 ), and both studies aim to boost the welfare of all advertisers on search engine platforms. More recently, a new stream of work has emerged which formulate the SEM bidding optimization as a dynamic pricing problem (Cheng, 2018 ) by incorporating the sequential behavior of SEM ads (Dayanik and Parlar, 2013 ; Shen et al, 2020 ).…”

Optimal dynamic multi-keyword bidding policy of an advertiser in search-based advertising

“…In the current paper, we consider a stochastic model for this environment. Extending the single keyword formulation of Dayanik and Parlar (2013), we study the problem of finding a bidding policy for multiple keywords in a general auction setting as a continuous-time optimization problem, and we solve it for an optimal policy using stochastic dynamic programming. This is the main contribution of the paper.…”

“…For a single keyword, the first optimal dynamic bidding strategy in a stochastic model with closed bids and under strict budget constraint was derived by the aforementioned paper Dayanik and Parlar (2013). The straightforward generalization of that model to K -many keywords with a given budget B over a planning horizon T would require solving the problem sup…”

“…We char-acterize the value function as the unique solution of a differential equation on this two-dimensional space and show how to numerically obtain it; see (34). This approach shields us from the computational issues and suboptimality of the straightforward extension of Dayanik and Parlar (2013) described above. To highlight the importance of using such an approach, in Sect.…”

“…To highlight the importance of using such an approach, in Sect. 6, we numerically compare the performance of our multi-keyword optimal policy to a number of heuristics in which the budget is split for each keyword and one-keyword policy of Dayanik and Parlar (2013) is applied separately. These heuristics differ in the way they partition the budget for different keywords.…”