Article Preview
TopIntroduction
Keyword segmentation and budget allocation are interdependent and complex decisions in search advertising (Ayanso & Mokaya, 2013; Ayanso & Karimi, 2015; Jansen & Spink, 2009). Despite many studies on mechanism design and bidding strategies for keyword auctions (Vragov and Shang, 2020; Vragov et. al, 2019), there has been limited research on the ad budget allocation problem. Search advertisers must decide which keywords to bid on, how to organize ad campaigns, and how much they should spend across planning horizons, campaigns, and individual keywords (Zhang & Feng, 2011; Yang et al., 2012; Yang et al., 2014). With millions of dollars being spent daily on search advertising, efficient allocation of the budget is of great importance.
This study attempts to bridge the research gaps around the issues of keyword segmentation, campaign organization and budget allocation in the search advertising body of knowledge. The study addresses these issues by leveraging an existing marketing framework as a theoretical lens and simulation modelling as a method of capturing this complex process. It examines whether advertisers can improve their budgeting decisions by employing keyword segmentation and performance-based budget allocation strategies. Specifically, using the buying funnel model (Jansen & Schuster, 2011) as the basis of keyword segmentation and campaign organization, this study examines Cost-based, Volume-based, and Clicks-based budget allocation strategies and evaluates the performance implications for firms with different product/service offerings.
Anecdotal evidence from search advertising practices shows that most advertisers manage a large number of accounts, campaigns, and keywords that vary in performance, budget consumption, and advertising goal orientation. To remain competitive, firms need to know the various decisions involved in search engine advertising as well as competitors’ campaign strategies in both organic (natural) and paid search advertising (Ayanso & Karimi, 2015; Jansen & Mullen, 2008). The absence of a systematic framework and decision support tools for many of the tasks involved in this process could lead to arbitrary decisions and resource inefficiencies. In addition, measuring keyword-level performance is important because paid search advertising operates at the keyword level (Özlük & Cholette, 2007; Rutz, Bucklin, & Sonnier, 2012). Given the significant amount of money spent on keyword advertising, marketers need broader insights into this process (Dhar & Ghose, 2010; Jansen, Sobel, & Zhang, 2011; Lu & Zhao, 2014). Effective keyword management requires identifying and creating different categories of keywords for improved budget utilization. Therefore, from a theoretical view, this study offers a unique insight into the budget allocation problem by leveraging the buying funnel model as the theoretical foundation, as well as integrating the campaign budget allocation decision with keyword segmentation.
Among the specific challenges advertisers face is the volatility in search demand which may have a direct effect on the performance of campaigns by causing budgets to run out early. In order to better manage their productive keywords, advertisers need to measure performance on a continuous basis and assess their impact on budget utilization. Therefore, from a practical perspective, this study provides insights into operational issues related to budget utilization as well as keyword categorizations that align with campaign strategies and objectives.
The rest of this paper is organized as follows. In the next section, we provide a review of the relevant literature in search engine advertising (SEA) and highlight the theoretical foundations. The following section presents the problem formulation and decision scenarios, along with details of the simulation model. Then, the experimental setting, the data and performance metrics used in this study are presented, followed by the simulation results. After discussing the results and their implications, the paper concludes and outlines the study’s limitations and future research directions.