Excerpt
TABLE OF CONTENTS
1. INTRODUCTION
2. CONCEPTUAL FRAMEWORK
3. METHODOLOGY
3.1. Database Search Query
3.2. Summary and Categorization
4. EMPIRICAL FINDINGS
4.1. Analysis and Classification of Research Articles
4.2. Search Engines
4.2.1. Search Engine Usefulness, Utilization, and User Experience
4.2.2. Search Engine Profitability
4.2.3. Type of Advertisement Auctions
4.2.4. Organic Page Rank
4.3. Advertisers
4.3.1. Paid Search Profitability and Effectiveness
4.3.2. Bidding Strategies and Keyword Competition
4.3.3. Ad Position Performance
4.4. Consumers
5. DISCUSSION
5.1. Theoretical Contribution
5.2. Practical Contribution
5.3. Limitations
5.4. Future Research Implications
6. CONCLUSION
REFERENCES
APPENDIX
List of Figures
Figure 1. Conceptual Framework of Search Engine Research
Figure 2. Number of Relevant Articles per Year
Figure 3. Number of Relevant Articles per Topic
List of Abbreviations
Abbildung in dieser Leseprobe nicht enthalten
Abstract:
This paper examines the academic research that has been published during the past 20 years concerning search engine marketing. It classifies the most prominent papers and establishes a conceptual framework consisting of three broad areas and 20 different research themes. This is the first literature review of search-engine-related research, proposing a comprehensive conceptual framework.
Purpose:
The purpose of this study is to conceptualize past research on search engines, develop a framework of topic areas and their relationships among each other, and identify research gaps and questions for future research.
Research Design:
The literature review covers all academic research papers from five acclaimed journals of the past 20 years that match predefined keywords and deal with search engines as parameters in modern-day marketing.
Originality and Value:
This literature review of search engine research is the first of its kind, creating a framework of research fields exclusively for online search. Furthermore, it presents the related academic work in an interdependent structure of topics and suggests future research fields.
Keywords: search engine, paid search, keyword auctions, digital marketing, online marketing
1. INTRODUCTION
Digital marketing, social media marketing and mobile marketing are an important part of virtually every business’ marketing strategy. For the time period from 1987 to 2015, researchers have carried out literature reviews (Lamberton & Stephen, 2016; Ngai, 2003; Schibrowsky, Peltier, & Nill, 2007) that gave overviews of the evolution of digital marketing research as well as current trend topic areas. It is remarkable that until now no literature review has focused on the topic of search engine marketing alone, given its prominent role in modern-day online business practice. The prevailing relevance of search engine marketing can be witnessed in the European Commission’s recent decision in July 2018 to impose a record antitrust fine of €4.34 billion on Google for abusing its market power (Satariano & Nicas, 2018). As the owner of the market leading Smartphone operating system Android, “Google has used Android as a vehicle to cement the dominance of its search engine”, as stated by competition policy commissioner Margrethe Vestager (European Commission, 2018). The previous record antitrust fine of €2.4 billion in June 2017 was imposed on Google as well – also related to unfair search engine competition (Scott, 2018). The European Commission’s rigorous measures against the world’s leading search engine provider and the record-breaking amount of the fines, which were calculated based on Google’s paid search revenue from Android devices in the European Economic Area, show the enormous influence that search engines have on electronic commerce today (European Commission, 2018).
Search engines’ major impact on not only businesses but also economies necessitates a comprehensive overview of past literature.
The latest literature review in the field of digital, social media, and mobile marketing was carried out by Lamberton and Stephen (2016), who built onto the two previous internet marketing literature reviews conducted by Ngai (2003) and Schibrowsky et al. (2007). They noticed that despite the importance and long existence of search engines as a research field, relatively few citations in economics literature can be observed. For example, the topic “search engines” only showed 3.4 mean cites per year from 2001 to 2015.
They further argue that despite a large body of research, no comprehensive answers have been researched to fundamental questions such as “What is the optimal balance between online and offline marketing?” or “What makes a digital marketing initiative a success for firms or consumers? Are there metrics beyond ROI that matter?” (Lamberton & Stephen, 2016, p. 166); questions that should also be considered in research about search engine marketing.
It is particularly interesting to examine how the different research topics, which were identified as emerging digital marketing research topics by Lamberton and Stephen (2016), are covered in search engine articles. These topics included collective behavior, online and offline crossover, as well as regulation and digital privacy issues. Additionally, it needs to be discussed if the main categories that Schibrowsky et al. (2007) found will also be applicable to search engine research: Consumer behavior, business to business, strategy, communication, product/brand, distribution, pricing, research issues, political legal, education. It seems that these categories are too general to create additional value for a search-engine-specific research framework.
This leads to this review’s guiding research question: Which elements of search engine advertising play a role in academic research and what is their relationship to each other? As part of this investigation, it should also be considered which aspects have been studied extensively and which areas should be examined more thoroughly.
Hence, the purpose of this work is not to provide a comprehensive literature review of every article related to search engines, as Lamberton and Stephen (2016) already pointed out. The scope of this work is simply too limited. The real contribution will be the identification of common research areas, labelling of search engine research fields and uncovering the underlying connections between the research topics.
Future research will benefit from this structured overview of academic search engine research. It will make it easier to categorize and locate a research article within the search engine area. Blind spots in the collected past literature are revealed and suggested for future studies.
The methodological approach is similar Ngai’s (2003) work, where he focused his search for relevant internet marketing literature on three publication areas, namely marketing, general business management, and information technology. He sorted the resulting articles into broad categories and then segmented them into more specific subcategories. Subsequently, he analyzed the distribution of the years of publication and the distribution of articles by subject.
The difference in this literature review is that mostly marketing journals were considered, due to the scope of this work and the slight results in the business management area.
2. CONCEPTUAL FRAMEWORK
Past research of search engine marketing involves at least one of the following three market participants, from now on referred to as agents: search engines, advertisers, consumers.
The analysis of the 52 articles that matched the search query resulted in the identification of 20 different research subtopics. The subtopics are pictured as rectangular boxes (see Figure 1. Conceptual Framework of Search Engine Research, p. 10). These research subtopics can be collocated according to their connection to search engines, advertisers or consumers.
Half of the subtopics could be distinctlyassigned to one of the agents, for example “Paid Search Profitability”. Although “Paid Search Profitability” is obviously influenced by consumers and search engines as well, it is primarily dependent on the advertisers’ marketing mix, advertising budget, search engine advertising strategies, internal cost accounting and many other factors. This logic was applied to all subtopics: Whenever a topic was primarily dependent on a single agent, it was assigned to that specific area. In the framework illustration, the agent areas are all connected with double-headed arrows to emphasize their interdependence.
The remaining ten subtopics that could not be attributed to one of the three agents were positioned in between two agents, illustrated as a grey-shaded area. All subtopics in one grey area have in common that they are heavily dependent on not one but two agents. “Paid Search Effectiveness”, as an example, cannot be explained by solely analyzing advertisers or search engines. Each agent contributes their part to the effectiveness; advertisers might need to create engaging ads, and search engines should display them to the right people.
This visualization uncovers the overall emphasis on advertiser-related marketing research and simultaneously highlights the lack of consumer-centric research. Needless to say, academic marketing research has focused on a plethora of consumer-centered topics. However, in this context, consumer-centric research means topics that are search-engine-related and analyze the consumer perspective of it.
Naturally, advertisers and consumers are strongly connected, and all advertiser subtopics involve consumers as well. However, purely consumer-oriented matters have rarely been considered. Topics directly related to consumers were consumer trust in search engine results or advertisements, and consumers’ price sensitivity.
The presented conceptual framework of search engine research is the first attempt of structuring this academic research area. It raises no claim to completeness; the arranged topics were merely a result of this literature review. It is certainly possible that other topics related to search engines have been studied but were not included in this framework’s underlying sample of 52 articles.
Abbildung in dieser Leseprobe nicht enthalten
3. METHODOLOGY
In order to understand which academic area exhibits most of the research efforts for search engine related topics, journals in the fields of entrepreneurship, marketing, and general business management were examined. Other research fields such as computer science and information system will most likely have done research on the highly technical subject of online search engines. However, as already argued by Lamberton and Stephen (2016), these academic fields are beyond the scope of this literature review’s objective, which is primarily business-related.
Journals of the two best rating categories in the VHB-JOURQUAL 3 ranking were included in this preliminary search. VHB-JOURQUAL is a rating list of academic business journals compiled by Verband der Hochschullehrer für Betriebswirtschaft (VHB), a German organization of business studies professors.
VHB rates journals with letters from A+ to D. Their final ratings are derived from the median of all ratings submitted by VHB’s more than 1,100 academic members. An A+ rating signifies an outstanding, world-leading academic business journal. Journals given an A-rating belong to the leading ones within academic business disciplines. B-rated journals are important and reputable academic publications. Journals with a C-rating are recognized academic business publications; a D-rating signifies a general journal in the academic business area without special recognition.
As a temporal limitation, only publications of the past twenty years were considered. The intention was to make sure that all years since the foundation of today’s leading search engine Google in 1998 were covered. It would have been particularly interesting to compare the research in this field before and after the market entry of this disrupting company. However, as in Lamberton and Stephen (2016) or Schibrowsky et al. (2007), the time restriction had to be made due to the scope of work. In order to cover the full twenty years, all articles published from 1997 to 2017 were searched and in addition the year 2018 until the month of July.
The keywords that were used to filter the database search results were an important tool to efficiently identify related publications. The selected keywords were quite generic search-engine-terms and could appear in the title, the abstract, or a paper’s keyword section: Search Engine, Paid Search, Sponsored Search, Google, Yahoo, Bing, SEO, Search Advertising, Search Marketing. This enabled a very broad search that included every paper that was remotely connected to a search-engine-related topic.
The entrepreneurship rating list showed four A-rated and eight B-rated journals (VHB, 2015b). Out of these twelve top-rated journals in the field of entrepreneurship, only five journals, two A-rated and three B-rated journals, published a total of eight articles that matched the search query for the defined time frame. Interestingly, when the four VHB-top-rated journals for entrepreneurship Research Policy, Journal of Business Venturing, Entrepreneurship Theory and Practice, and Strategic Entrepreneurship Journal were searched, only five matching articles could be found. Furthermore, after conducting a preliminary analysis of each article’s abstract, none of the matching articles were relevant for the search engine literature review.
The VHB rating list for general business management featured eight A+-rated and nine A-rated journals (VHB, 2015a). Of that list, five A+-rated journals published thirty articles that matched the search query, whereas four A-rated journals published four potentially relevant articles. The abstract analysis showed that sixteen of the thirty-four articles were relevant for a literature review.
The VHB rating for marketing journals showed a total of four A+-rated and seven A-rated journals (VHB, 2015c). The four A+-rated journals published sixty research articles that matched the search query in the time frame while the A-rated journals published eleven. The first abstract analysis resulted in forty-three relevant articles from A+-rated journals and seven relevant articles from A-rated journals.
This confirms the finding of the last significant digital marketing literature review by Lamberton and Stephen (2016) that digital marketing is predominantly researched in the academic discipline of marketing rather than in other business disciplines.
After this initial scan of reputable business journals, the final database search query was decided, and the results were summarized and categorized, which will be explained in the following.
3.1. Database Search Query
In order to limit the search for academic journals to the ones that were used by Lamberton and Stephen (2016), only articles from Journal of Marketing Research, Management Science, Journal of Marketing, Journal of Consumer Research, and Marketing Science were taken into consideration. These are the journals that were given an A+-rating in the VHB-JOURQUAL 3 marketing rating and the A+-rated journal Management Science from the general business management rating list. This makes this literature review to a certain extent comparable with the literature review by Lamberton and Stephen (2016), who analyzed the exact same set of journals. As mentioned before, other journals with lower ratings in the VHB-JOURQUAL 3 list also published articles with relevancy to search engine topics, but the scope of this literature review had to be limited for practical reasons.
The database search query included the aforementioned topic-related keywords, the five selected A+-rated journals, and the time frame of more than 20 years from 1997 to mid 2018 (July). The database of Business Source Premier, as in Schibrowsky et al. (2007), was accessed through the Web of Science interface1 to search and filter relevant papers. The search resulted in 78 articles that matched the query. The abstract analysis resulted in 26 articles that were not relevant, leaving 52 articles that researched search-engine-related topics.
The irrelevant articles were mostly mismatched, because they used search engines as a part of their studies’ methodology. Oftentimes, search engines are used in field experiments or historical search data is analyzed to study an unrelated topic. Sometimes, articles mismatched because the abstract contained one of the brand names, e.g. Yahoo, Google, or Bing, but just referred to them as an example and did not actually research a topic related to search engines.
3.2. Summary and Categorization
In order to determine the primary topic area of each article, the content of each article was analyzed. The analysis showed that 26 articles were not relevant for this literature review, e.g. because the research merely used search engines as a tool but not as their main research object. Information about each study’s sample, research method, and key findings were summarized for each of the remaining 52 articles, subsequently assigning them to relevant research topic categories (see the literature summary table in APPENDIX). At first, detailed research topic descriptions were given to each article. Later, these detailed topic descriptions were merged into broader topic categories. The topic categories were necessary to create the overarching conceptual research framework that structures all research connected to search engines. The framework aids in categorizing existing and future research as well as in uncovering blind spots, which could be potential future research areas. The literature table also contains the scientific method that was used, classified into quantitative as well as qualitative research or economic model as research method. The economic model was specified as quantitative when the study used empirical data to test the model, and as qualitative if only theoretical economic modelling was applied.
4. EMPIRICAL FINDINGS
4.1. Analysis and Classification of Research Articles
Abbildung in dieser Leseprobe nicht enthalten
Figure 2. Number of Relevant Articles per Year
Before the content analysis of each article was carried out it was helpful to create a diagram that showed the number of articles published per year (Figure 2). This made it possible to understand in which time periods of the late 20th and early 21st century most of the research was conducted. Figure 2 reveals that most articles have been published since 2008. Only three articles were published in the ten years from 1997 to 2007. The year with most publications was 2011 with eleven articles, followed by 2014 and 2015 with nine articles each.
The analysis of all 52 academic research articles showed that almost all studies touched multiple different topics connected to search engines. Figure 3 illustrates the popularity of certain topics and exposes the topics that were only touched a few times over 20 years.
Research articles that only dealt with a single topic were about search engine usefulness, organic page rank, paid search effectiveness, bidding strategies, and search engine utilization (Bradlow & Schmittlein, 2000; Danaher & Dagger, 2013; Du, Hu, & Damangir, 2015; Katona & Sarvary, 2008; Rutz, Sonnier, & Trusov, 2017; Skiera & Abou Nabout, 2013).
The most frequently researched topics were paid search effectiveness (23 articles), search engine profitability (15 articles), ad position performance (14 articles), and search engine utilization (14 articles). Topics that were investigated in the least amount of articles were attribution models (H. (Alice) Li, Kannan, Viswanathan, & Pani, 2016; H. Li & Kannan, 2014), search engine usefulness (Bradlow & Schmittlein, 2000; Chesnes, Dai, & Zhe Jin, 2017) and click fraud (Wilbur & Zhu, 2009).
Abbildung in dieser Leseprobe nicht enthalten
Figure 3. Number of Relevant Articles per Topic
4.2. Search Engines
4.2.1. Search Engine Usefulness, Utilization, and User Experience
Before search engines became an essential part of everyone’s online navigation, early studies in the field were concerned with the usefulness of search engines. Bradlow and Schmittlein (2000) searched 20 marketing phrases with several popular search engines in order to evaluate how well they covered the world wide web. This was at an early stage of search engine development and today’s market leader Google was not even included in the study. Bradlow and Schmittlein (2000) found that search phrase and URL characteristics significantly affect search engine outcomes. They already understood that the total number of web pages indexed by a search engine is the biggest factor for search effectiveness. A common finding across engines was that the more links were included on a web page, the more likely that web page would be found. Because the indexing was not extensive, users needed to use multiple engines to achieve good results.
The only other study in the sample that was concerned with search engine usefulness researched the effect of prohibiting foreign pharmacies to buy sponsored search ads (Chesnes et al., 2017). Certain foreign pharmacies that were not certified by the National Association of Boards of Pharmacy (NABP) were banned from paid search ads. Chesnes et al. (2017) found that banned pharmacies without any other certification experienced much less clicks than before the ban. On the search results page for a certain drug, the search engines displayed general drug information, such as side effects, usage advice or warnings, instead of the sponsored links. This increased usefulness of search engine results for consumers could also be seen in the decreased link clicks on organic listing of banned pharmacies.
At the beginning of the rise of search engines, it was also studied how users utilize the search services to navigate the internet. Telang, Boatwright and Mukhopadhyay (2004) studied the internet-use behavior from 1998 to 1999 and found that users tend to decrease their search engine usage over time. They also achieved a precise prediction of visit timing with a mixture model that explicitly incorporates user schedules. Nonsearch features like email or news tend to increase the amount of search engine visits and prevent users from reducing their search engine use, leading to a regular and continued use.
Concerning consumers’ utilization of sponsored ads, Yao and Mela (2011) found that around 10% of search engine users are responsible for 90% of the clicks on sponsored search results. The consumers that click the most also give greater value to the ad position, i.e. top ad positions are clicked more often by people who click a lot of ads. Once the consumer has clicked the ad and found the website that is useful for them, they generally return by typing in the website address directly (Rutz, Trusov, & Bucklin, 2011). The notion that two different user groups for search engines exist, high-involvement and low-involvement consumers, was also supported by Jerath, Ma and Park (2014) and Chan and Park (2015). According to Jerath et al. (2014), high-involvement consumers typically search for less popular keywords and show more (sponsored) clicks per search compared to low-involvement consumers, who search for popular keywords. Edelman and Lai (2016) on the other hand divide search engine users into searchers who consider conspicuous search results (salience searchers) and searchers who assess a search result’s relevance (relevance searchers).
Particularly interesting is Hu, Du, and Damangir's (2014) finding that the interests reflected in Google users’ search queries may be considered as representative of the general population. In connection with Joo, Wilbur, Cowgill and Zhu (2014), it shows that search engine utilization can reflect society’s behavioral patterns. Joo et al. (2014) found that advertisements on television increase the advertised brand’s share of keywords searched and also increases the number of searches in the relevant product category. Hence, it is not surprising that product feature search trends may be treated as reflective indicators of feature importance trends (Du et al., 2015). These insights have managerial implications that go beyond sales forecasting but can also be used in product design, budget allocation, advertising design and budgets, and product and inventory planning (Du et al., 2015).
The user experience seems to have a great impact on search engine utilization and also on search engine profitability. As mentioned before, search engines use additional services and features, e.g. email services, news services, or file storage, as means for user retention (Telang et al., 2004). At the same time, removing filtering and sorting features of the search engine would decrease its usability and would cause a 2.9% reduction of search engine revenue (Yao & Mela, 2011). User experience is therefore directly related to search engine profitability.
Ironically, providing a good user experience can also hurt search engines’ revenues: High-quality websites that are ranked higher organically provide a great user experience, as it simplifies the search for the best result. This attracts more consumers to the search engine but hurts its revenues because the high-quality websites buy fewer paid search results (Berman & Katona, 2013). Nevertheless, when using the customer-utility-based ranking model to rank paid search ads, the highest search engine revenue occurs (Ghose, Ipeirotis, & Li, 2014).
4.2.2. Search Engine Profitability
Chen, Liu and Whinston (2009) researched the fundamental question of search engine advertisement, which is the share structure problem, i.e. how many advertising resources to give to the highest and lowest bidder. According to the model, the share structure problem can be solved in dependency on advertisers’ price elasticity, return factors of different shares, and changes in total advertising resources. In today’s search engine advertising programs, these parameters are called daily budget, click-through-rate (CTR), and the available space on different devices, such as smartphones, tablets, or desktop computers.
[...]
1 Website Address: http://www.webofknowledge.com