No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Identification of influencers — Measuring influence in customer networks
Abstract
Viral marketing refers to marketing techniques that use social networks to produce increases in brand awareness through self-replicating viral diffusion of messages, analogous to the spread of pathological and computer viruses. The idea has successfully been used by marketers to reach a large number of customers rapidly. If data about the customer network is available, centrality measures provide a structural measure that can be used in decision support systems to select influencers and spread viral marketing campaigns in a customer network. Usage stimulation and churn management are examples of DSS applications, where centrality of customers does play a role. The literature on network theory describes a large number of such centrality measures. A critical question is which of these measures is best to select an initial set of customers for a marketing campaign, in order to achieve a maximum dissemination of messages. In this paper, we present the results of computational experiments based on call data from a telecom company to compare different centrality measures for the diffusion of marketing messages. We found a significant lift when using central customers in message diffusion, but also found differences in the various centrality measures depending on the underlying network topology and diffusion process.
... Within the marketplace contexts, scholars have used celebrity endorsement theory , communicative ecology theory ( Seol et al., 2016 ), flow theory , identification theory ( Yoo et al., 2013 ), motivational theory ( Yoo et al., 2013 ), network theory ( Kiss & Bichler, 2008 ), product involvement theory , psychological theory , social presence theory ( Xu et al., 2012 ), social support theory ( Balaji et al., 2016 ), stakeholder theory and uses and gratifications theory Xu et al., 2012 ) for increasing their brand influence on present and potential customers. Scholars had used crowd wisdom theory ( Eickhoff & Muntermann, 2016 ) and information processing theory for extracting the market intelligence from user generated content present on social media. ...
... Bitter and Grabner-Kräuter (2016) had illustrated an online environment in which they had posted different variations of a brand-related post on Facebook and had triggered different responses from users. Kiss and Bichler (2008) had stated selection of an initial set of customers for a marketing campaign matters for maximizing the reach of the messages. and Li and Lai (2014) had purposed the mechanism for summarizing collective opinions present on social media. ...
... Within algorithmically computed content topic models and sentiments of the posts had been studied ( Eickhoff & Muntermann, 2016 ;He et al., 2015 ;Jiang et al., 2014 ;Li & Lai, 2014 ;Rui et al., 2013 ;Yu et al., 2013 ). Within the network parameters following characteristics had been studied: "in-degree " or "out degree " of connections ( Kazienko et al., 2013 ;Phang et al., 2013 ;Schlereth et al., 2013 ;Susarla et al., 2012 ); position within network ( Kazienko et al., 2013 ;Levina & Arriaga, 2014 ); network size ( Schlereth et al., 2013 ;Seol et al., 2016 ); betweenness, closeness and reciprocity ( Phang et al., 2013 ); tie strength ( Balaji et al., 2016 ;Bitter & Grabner-Kräuter, 2016 ;Li & Du, 2011 ;Li & Lai, 2014 ;Luarn et al., 2015 ); homophily ( Li & Du, 2011 ;Pagani & Mirabello, 2011 ;Susarla et al., 2012 ); centrality ( Kiss & Bichler, 2008 ); and social influence ( Levina & Arriaga, 2014 ;Li & Shiu, 2012 ;Lin et al., 2015 ). Within the platform characteristics followers ( Piehler et al., 2019 ;Zhang et al., 2011 ) and likes ( Chang et al., 2019 ;Lee et al., 2015 ;Wessel et al., 2016 ) had been studied. ...
Every aspect of human activities has been influenced by social media, but how this influence is affecting individual decision making in different context had not been studied yet. This article tries to investigate social media influences at individual level with respect to different contexts such as organization, marketplace, and social environment. 132 articles had been selected for the review process. The findings have been presented using the lens of Theory, Context, Characteristics and Methodology. This article presents social media influence phenomenon within an individual. The article concludes by highlighting literature gaps and future research directions. This review makes the relevant contribution to the field of computer mediated communications.
... The field has seen increased research interest over the past few decades, ostensibly due to the rise of online social media platforms and e-commerce. There are numerous problems of interest including whether, or how quickly, people's opinions converge to agreement [2,3], how to identify the most influential actors [4], and the effects that stubborn and zealous actors have on the system dynamics [5,6,7]. One of the important areas that is frequently overlooked is that of 'opinion diversity'. ...
... We also demonstrate how, depending on how the underlying distribution is defined, random fluctuations can capture both endogenous or exogenous factors that contribute to opinion formation. 4. We empirically validate the predictions of the noisy DeGroot model by testing the extent to which it captures variations in opinions in online news data. ...
... In contrast, for the dynamics of Eqs. (4) and (5) to be stable and stationary, the adjacency matrix must be strictly substochastic, i.e., N j=1 Ai,j < 1. ...
We investigate the impact of noise and topology on opinion diversity in social networks. We do so by extending well-established models of opinion dynamics to a stochastic setting where agents are subject both to assimilative forces by their local social interactions, as well as to idiosyncratic factors preventing their population from reaching consensus. We model the latter to account for both scenarios where noise is entirely exogenous to peer influence and cases where it is instead endogenous, arising from the agents’ desire to maintain some uniqueness in their opinions. We derive a general analytical expression for opinion diversity, which holds for any network and depends on the network’s topology through its spectral properties alone. Using this expression, we find that opinion diversity decreases as communities and clusters are broken down. We test our predictions against data describing empirical influence networks between major news outlets and find that incorporating our measure in linear models for the sentiment expressed by such sources on a variety of topics yields a notable improvement in terms of explanatory power.
... Of the network structural variables, degree (first-order) centrality indicating the number of direct ties an alter holds is often used to reflect the person's interconnectedness or sociability. Along with degree centrality, sociologists, and network scientists have long explored the possibility of developing alternative centrality measures, especially for indirect (higher-order) relations (Freeman, 1979;Kiss & Bichler, 2008). Among many others, eigenvector and Katz centrality have been widely used, but known to have a potential problem, in that, a wellconnected node passes centrality to all of its contacts, thereby artificially inflating the centrality of even those at marginal positions (i.e., not everyone connected to a well-known person is also well-known). ...
... With PRC, the centrality passed to a node i from a node j is diluted depending on the number of connections of the node j. This means that being tied or merely following a prominent node with many connections would not significantly increase one's PRC (Kiss & Bichler, 2008). This also suggests that when an alter is tied to some well-connected others outside of a local network, excluding such external ties might not yield substantial changes in the person's PRC, which is beneficial for the studies dealing with partial networks. ...
Most messages on social media platforms are reportedly posted by a small number of active communicators, while the great majority of users remain silent as lurkers who read but seldom write. Despite extensive research to date, it remains unclear why such a disparity in individuals’ participation in social media exists. Drawing on the behavioral data of 15,633 Facebook users nested in 73 local networks, this study attempted to examine how the structural properties of networks give rise to the highly skewed distribution of message contributions between individual users. Multilevel statistical analyses of the data revealed that the participation disparity among individuals might be in part a function of the structural characteristics of networks in which they are embedded, suggesting that being active or silent in the social media environment is largely conditional on the surrounding network structures.
... In network science, this is done by using a centrality measure which assigns nodes a centrality value with larger values indicating greater importance. Centrality has been developed over the last 70 years [2][3][4][5][6][7][8][9][10][11][12][13][14][15] and is a core part of many introductory texts on Network Science [16][17][18][19] . The simplest centrality measure is a degree, the number of neighbours each node has, probably the first node property examined in any study. ...
... Closeness centrality measures the impact of an author on a field and their social capital 10 . When used to select potential leads in customer data, closeness led to a significant gain in the success rate 9 . In an air transport network, it has been shown that the closeness of a city is highly correlated with socio-economic indicators such as gross regional domestic product 13 . ...
Measuring the importance of nodes in a network with a centrality measure is an core task in any network application. There many measures available and it is speculated that many encode similar information. We give an explicit non-linear relationship between two of the most popular measures of node centrality: degree and closeness. Based on a shortest-path tree approximation, we give an analytic derivation that shows the inverse of closeness is linearly dependent on the logarithm of degree. We show that our hypothesis works well for a range of networks produced from stochastic network models and for networks derived from 130 real-world data sets. We connect our results with previous results for other network distance scales such as average distance. Our results imply that measuring closeness is broadly redundant unless our relationship is used to remove the dependence on degree from closeness. The success of our relationship suggests that most networks can be approximated by shortest-path spanning trees which are all statistically similar two or more steps away from their root nodes.
... The links between student degree centrality and performance have been the object of many student network studies (e.g., (Thomas, 2000); (Yang and Tang, 2003); (Russo and Koesten, 2005); (Cho et al., 2007); (Obadi et al., 2010); (Hommes et al., 2012); (Gašević et al., 2013); (Mushtaq et al., 2016); (Zwolak et al., 2017); (Vargas et al., 2018); (Liu et al., 2018); (Saqr et al., 2018); (Vignery and Laurier, 2020a)). Additionally, many centrality measures -which each reflect dimensions of centrality that are different -were developed in SNA to assess node centrality within a graph ( (Kiss and Bichler, 2008;Landherr et al., 2010;Song et al., 2015); (Ghazzali and Ouellet, 2017); (Ashtiani et al., 2018)). However, even if there are counterexamples (e.g., (Zwolak et al., 2017)), most of the studies dedicated to student networks have conceptualized student centrality as degree, closeness and/or betweenness centralities (Vignery and Laurier, 2020b), without investigating other dimensions than those covered by these three measures. ...
... Part of the Hyperlink-Induced Topic Search (i.e., HITS) algorithm, and developed by , this measure was initially developed to assess web pages. Authorities are pages possessing many incoming links and containing high-quality information ( (Kiss and Bichler, 2008;Fouss et al., 2016;Lü et al., 2016)). Related to student networks, the authority score reflects students' importance according to the number of important nodes -i.e., hubs -that point towards these students. ...
The impact of student networks on academic performance has gained importance as a research subject. In addition to well-known centrality measures (i.e., degree and closeness centralities), this study tests indices that received less attention to predict student performance. This set of measures allows for distinguishing the effects on performance of being connected, being located in advantageous position(s), being connected to central peers and being located within connected neighborhoods. Besides, studies on links between network density and student achievement are rare. This research investigates the combined impacts of student centrality and network density on academic performance. We asked 574 college students about their friendships, and drew the network from the collected information. We used the Exponential Random Graph Models to impute missing friendship ties. Then, we applied a hierarchical clustering approach that identified sub-communities within the student network and we computed the density within each sub-community to study density. Finally, we used hierarchical modeling to predict student performance, i.e., by centrality at the student level and by density at the network level. Results demonstrate a positive impact of the geodesic k-path and of the closeness centralities on GPA, together with a positive impact of cluster’s density on performance, which seems, however, bounded by a ceiling effect.
... A popular approach is to use an individual's centrality within a social network. Centrality, which is defined in more detail in section 2.4.4.3, is a measure of how influencial someone is based on where they are positioned within a social network [66,64,67,68]. Other measures ignore centrality. ...
... The field has seen increased research interest over the past few decades, ostensibly due to the rise of online social media platforms and ecommerce. There are numerous problems of interest including whether, or how quickly, people's opinions converge to agreement [44,140], how to identify the most influential actors [66], and the effects that stubborn and zealous actors have on the system dynamics [141,142,35]. One of the important areas that is frequently overlooked is that of opinion diversity. ...
The coexistence of diverse opinions is necessary for a pluralistic society in which people can confront ideas and make informed choices. The media functions as a primary source of information, and diversity across news sources in the media forms the basis for wider discourse in the public. However, due to numerous economic and social pressures, news sources frequently co-orient their content through what is known as intermedia agenda-setting. Past research on the subject has examined relationships between individual news sources. However, to understand emergent behaviour such as opinion diversity, we cannot simply analyse individual relationships in isolation, but instead need to view the media as a complex system of many interacting entities. The aim of this thesis is to develop and empirically test a method for understanding the network effects that intermedia agenda-setting has on the diversity of expressed opinions within the media. Utilising latent signals extracted from news articles, we put forward a methodology for inferring networks that capture how agendas propagate between news sources via the opinions they express on various topics. By applying this approach to a large dataset of news articles published by globally and locally prominent news organisations, we identify how the structure of intermedia networks is indicative of the level of opinion diversity across various topics. We then develop a theoretical model of opinion dynamics in noisy domains that is motivated by the empirical observations of intermedia agenda formation. From this, we derive a general analytical expression for opinion diversity that holds for any network and depends on the network's topology through its spectral properties alone. Finally, we validate the analytical expression in a linear model against empirical data. This thesis aids our understanding of how to model emergent behaviour of the media and promote diversity.
... First approach: Strategic Location (Social Network Analysis-SNA) Some researchers focused on the strategic location within a social network by considering its structural characteristics [9]- [11]. These structures resulting from connections among people are described as a set of nodes and undirected edges that connect pairs of nodes. ...
... Thus, some other studies found that "fringes" that are poorly connected nodes and characterized by low degree centrality might be particularly influential [14] and adoption may occur first at the fringes of the network instead of the hubs [15]. As mentioned in [11] hubs can spread viral information efficiently so it will be so effective to take hubs as a seed point. On the other hand, Closeness Centrality expands the diffusion of Degree Centrality on how close a node is to all other nodes in the network. ...
Online Social Networks (OSN) has become a major source of social data. The massive amount of data generated from OSN produce different information according to the type of analysis performed. The aim of this research is to improve the understanding of social influence and to identify influential users in OSN. The identification of influential users helps to improve marketing communication techniques. This research proposes a new framework for analyzing Online Social Network (OSN) data to identify influencers using Social Media Analysis and Data Mining techniques. Network centrality measures and text mining are used to analyze both structured and unstructured data.
... 723). "Influence" is a market research term for customer conversion (Keller & Berry, 2003) and "centrality measures" (Kiss & Bichler, 2008) used by commercial decision support systems to spread viral marketing campaigns. But from the point of view of users, "influence" is associated with numbers of friends, public sharing, personal appeal, popularity, and likeability (boyd, 2014). ...
... Marginalized actors who might initially seek to promote social relationships are rewarded with celebrification and career advancement according to the optimized, marketable taxonomies advertised on this site. In fact, although the term "influencer" refers to social media nodes tracked by commercial decision support systems (Kiss & Bichler, 2008), they have become socially synonymous with the positive "role models," in a cycle of virtue economics that treats representation as justice in itself, while preserving the upward distribution of profit. ...
This article examines Facebook’s role in the treatment of marginalized identity as currency. Recent examples of solidarity statements and corporate social responsibility rhetoric treat disenfranchised racial and gender identities as value-added competitive market quantities to boost brands. This trend also incentivizes marginalized actors to capitalize on their own disenfranchisement in pursuit of visibility and career advancement. The resulting identity politicking replaces communal care, grassroots social ties, solidarity, and interdependence with isolating market competition. This article diverges from scholars who trouble the differential value of identity—by troubling the valuation of identity itself. Facebook normalizes identity as private property in what I call a transition from identity politics to “personal identity economics.” I coin this concept and break it down into the following four factors: (1) The optimization of difference beginning in the 1970s, (2) Facebook’s algorithmic invasion of market logic into intimate aspects of life starting in the mid 2000s, (3) Ads Manager’s economization of identity into legible economic units, and (4) neoliberal corporate social responsibility rhetoric of “social good” as a profitable asset.
... Influence maximization is the main target of the TSS problem. When the structure of the target network is accessible, methods based on global centrality measures are a frequently-used method to select seeds [32], and numerous methods of centrality measure are proposed [33][34][35]. While the stochastic seeding strategy is appropriate for social networks with limited information [36]. ...
The high penetration of online communication in social networks provides a perfect context for launching a viral marketing campaign. Viral marketing strategies with adoption and promotion incentives are applied by companies for acquiring customers and occupying the market rapidly. In this paper, motivational thresholds and the SAN diffusion model are introduced to describe the massage diffusion process. Diffusion thresholds of dual incentives under homogeneous and heterogeneous networks are both deduced. We compare different effects of adoption and promotion incentives on marketing performances including the final penetration and the diffusion speed with numerical simulation. Simulation results show not only a greater influence of promotion incentives than adoption but also the necessity of adoption incentives for higher penetration. Besides, simulation results also show different diffusion characteristics between homogeneous and heterogenous networks, which provide management implications for selecting target customers with different network structures.
... The surveys were initially deployed via social-seeding through Facebook (Hinz et al., 2011). Specifically, the seeding included the use of strong ties (Watts, 2004) with the researcher and the targeting of influential members of the network (Kiss & Bichler, 2008). A social seeding strategy can be an effective way to quickly diffuse information through a network (Hinz et al., 2011) and, with regard to the present study, increase the sample size. ...
... among nodes rather than the intrinsic property of a node itself (Borgatti and Everett 2006;Kiss and Bichler 2008). A broader class of eigenvector centrality is also known as spectral centrality (Perra and Fortunato 2008), including the prominent Google's PageRank (Bryan and Leise 2006;Langville and Meyer 2006), the HITS algorithms (Kleinberg 1999), the Container Port Acessibility Index (Wang and Cullinane 2008), and the CPCI (Bartholdi et al. 2016). ...
This paper proposes a framework for evaluating the strategic importance of container ports based on their connectivity. The Container Port Connectivity Index is computed and decomposed into components according to the Liner Shipping Connectivity Index —each reflecting its contribution to the overall port importance score. The framework produces separate scores for each component, thus allowing port stakeholders to better comprehend why a particular port has become important, and for what reasons. The decomposition approach also allows more detailed analyses, and explanations of the impacts of major economic phenomena—i.e., the expansion of Panama Canal or the crumbling of Hanjin shipping—on the relative importance of ports within the Global Container Shipping Network , as more explanatory variables become available. Our computational results indicate that, while the connectivity of ports related to these events is impacted, changes on connectivity rankings could be adequately explained by the proposed decomposition scheme. The inbound connectivity of New York, for example, was slightly improved after the Panama Canal expansion—from the 29th place in Q1/2016 to the 24th place in Q2/2016—due mainly to the rise in the larger capacity of ships calling. However, in Q3/2016, its inbound rank returned to the 29th place, which was mainly due to the decline in the number of liner services available, number of liner companies, and number of ships calling. The effects of Hanjin’s bankruptcy, on the contrary, were more localized and relatively brief.
... Strategi marketing yang diciptakan oleh seorang influencer lebih banyak mengiklankan produk dengan menggunakan teknik dari mulut ke mulut. Menurut Bichler & Kiss (2008), efek pemasaran dari mulut ke mulut memiliki dampak positif terhadap kepuasan pelanggan. Pemasaran dari mulut ke mulut dan bentuk komunikasi yang dirangkai sendiri dari awal hingga akhir menjadi sebuah kelebihan yang dimiliki oleh seorang influencer media sosial. ...
Tingginya jumlah pengguna Instagram menyebabkan banyak produsen barang dan jasa memilih untuk mempromosikan produknya melalui jasa endorsement yang dilakukan oleh influencer media sosial. Strategi marketing menggunakan influencer ini dianggap penting karena dapat membantu produsen barang dan jasa untuk menciptakan citra positif produk dari mulut ke mulut serta dapat memenuhi tujuan bisnis antar kedua belah pihak. Berdasarkan fenomena tersebut, penelitian ini bertujuan untuk mengetahui strategi marketing yang digunakan oleh Influencer Arief Muhammad dan Tasya Farasya, serta melakukan komparasi jumlah likes dan comment dalam video endorsement yang diunggah di Instagram Feeds terhadap penggunaan tiga jenis strategi marketing, yaitu boasting of company, discount code, dan use of product, baik secara verbal maupun non verbal. Penelitian ini mengamati sebanyak 57 konten video endorsement yang diperoleh dari perhitungan rumus slovin, serta pengujian hipotesis dilakukan dengan metode analisis isi kuantitatif deskriptif dan statistik inferensial non-parametrik. Pengumpulan data pada penelitian ini dilakukan dengan mengamati unggahan video endorsement influencer terhadap penggunaan strategi marketing di Instagram Feeds yang kemudian dikodekan menggunakan coding book.
... In the last two decades, complex networks have attracted a great deal of attention since they efficiently describe a wide range of systems in biology, information technology, social science and so on [1]. Identifying important nodes in a network, such as super-spreaders of some disease in a population [2], influencers in a social network [3] and central cities in an air transportation network [4], is a fundamental task in network analysis. For this purpose, various centralities have been used so far [5]. ...
Betweenness centrality (BC) is a measure of the importance of a vertex in a graph, which is defined using the number of the shortest paths passing through the vertex. Brandes proposed an efficient algorithm for computing the BC scores of all vertices in a graph, which accumulates pair dependencies while traversing single-source shortest paths. Although this algorithm works well on static graphs, its direct application to dynamic graphs takes a huge amount of computation time because the BC scores must be computed from scratch every time the structure of graph changes. Therefore, various algorithms for updating the BC scores of all vertices have been developed so far. In this article, we propose a novel algorithm for updating the BC scores of all vertices in a graph upon deletion of a single edge. We also show the validity and efficiency of the proposed algorithm through theoretical analysis and experiments using various graphs obtained from synthetic and real networks.
... Voluntary transmission is the primary method through which viral marketing spreads. Kiss & Bichler (2008) describe "marketing approaches that employ social media to generate brand awareness all the way through" as "viral messaging and computer viruses that have the power to self-replicate." Kim & Lowrey (2010) have defined viral marketing as a marketing tactic or phenomena that promotes and inspires individuals to distribute marketing messages through social media networks among each other (Kim & Lowrey 2010). ...
It was the purpose of this survey to find out what people think about viral marketing. This study hypotheses that customers' opinions toward viral marketing are influenced by their perceptions of its utility and their own intrinsic motivations. Customers' perceptions of social media and viral marketing were also evaluated in this research. This study also demonstrated that customers' views of viral marketing are positively correlated with perceived incentives.
... Finding influential people in a science social network, e-mail network, discussion group, etc., assists with analyzing the networks [28,32]. For instance, a community is World Wide Web pages that are grouped according to their related subjects [17], functional modules such as cycle and paths in metabolic networks [22] or groups of people that are connected in social networks [31]. ...
... Depuis la fin des années 2000, la popularité et la diversité des réseaux sociaux en ligne dynamisent la recherche sur la détection des influenceurs par centralité. Beaucoup de travaux sont motivés par des applications en marketing, définissant les influenceurs comme les individus qui ont une forte capacité à promouvoir des produits dans les réseaux sociaux car ils ont un impact important sur la communication qui s'y déroule (Kiss & Bichler, 2008), (Heidemann et al., 2010). Il s'agit alors de trouver les mesures de centralité et les modélisations en graphe les plus aptes à identifier ces influenceurs. ...
Dans cette thèse, nous présentons la conception et l’évaluation d’un système pour détecter automatiquement les personnes influentes dans les médias sociaux, à partir des manifestations de leur action d’influence dans les communications interpersonnelles. Les approches pour la détection des influenceurs utilisent généralement, soit la structure de la communication entre les individus, soit l’analyse de son contenu. Le cadre théorique retenu dans notre thèse a la particularité de combiner ces deux types d’approches pour leur complémentarité. Nous caractérisons l’action des influenceurs à l’échelle d’un individu cible, depuis sa mise en œuvre jusqu’à ses effets, par des traits discursifs relevant aussi bien des messages envoyés par les influenceurs que de ceux envoyés par les individus influencés. La détection automatique de ces traits discursifs est faite avec des méthodes en traitement automatique des langues, basées sur des règles linguistiques et des modèles par apprentissage automatique. À l’échelle d’un groupe, l’action des influenceurs est caractérisée par leur position centrale dans un graphe social qui représente des actions interpersonnelles ayant eu cours à l’intérieur de ce groupe. L’hybridité de notre système consiste en l’utilisation des informations linguistiques sur les traits discursifs d’influence, extraits automatiquement depuis les messages textuels échangés entre individus, afin de construire les graphes sociaux.
... However, using WOM principles in a campaign presupposes that companies know clearly which are the most influential nodes in the social networks. Globally, discovering influential nodes from online social networks is one of the major avenues of WOM marketing research (Li et al., 2011;Duan et al., 2008;Kiss & Bichler, 2008). Therefore, identifying influential nodes on social media and the users who really have influence, is a big issue for companies who intend to do an influencer marketing campaign. ...
There are several paths in marketing to communicate with the consumers. More creative ways are reaching to the market increasing the fascia and trying to overcome customers' demands. Consumers have changed and are now more informed, demanding, and empowered. They are talking with brands and about them with other consumers but also about their personal experience, which means that the impact of the communication between them has evolved in massive terms. The new trends in digital communications has brought profound changes to the tourism sector. Tourists are now more critical in their decision-making process. The increasing access to new technologies by individuals has made travel research and planning easier, placing the stakeholders in a permanent challenge to meet the consumer's needs. In this respect, some points come out: are the new means of communications fundamental determinants in the consumer decision purchase in tourism products? Will companies beneficial to start including this new means as communication tools? Should they be incorporated in their communication plans?
... This connection places them before their social followers as their equals, generating empathy by sharing similar traits, such as age, culture, language, nationality, among others (Westenberg, 2016). Kiss and Bichler (2008) even pointed out that the impact of these «social influencers» is continuously increasing and that the attachment to the trends they generate in consumers gradually transforms social reactions at the time of purchase, a situation that is increasingly evident in our media reality. In this regard, Baella (2017) mentions the three moments of a traditional purchase and creates the «zero moment of truth» or ZMOT, which occurs when the user, feeling stimulated by an object, goes to the Internet to consult and make a final purchase decision. ...
The objective of this article is to analyze the contents and unconventional advertising narratives of the eight most important women in the world of fashion in Spain and Ecuador in relation to the number of followers and points of view, in order to identify the discursive and esthetic strategies and narratives that may reflect the keys to their experience as prescribers, through a content analysis based on the interpretation of the five most viewed videos between 2018 and 2019 from four Spanish to four Ecuadorian YouTube channels ( ME = 40) based on a three-round Delphi analysis sheet with a validity of W = 0.828 and α = .947. The content is analyzed from a qualitative perspective, which allows an in-depth exploration of the dimensions and indicators of impact and influence on YouTube channels. The research presents the findings that the influencers reviewed use crutches, idioms, and set phrases to identify with their audience. The audiovisual narrative is simple, maintaining its amateur style. Advertising positioning in the channels analyzed is given by identifying the brand in the spoken discourse, the presence of brand logos, advertisements and promotions, and the presence of products of the sponsoring brands.
... The massive adoption of social networks by users (especially professionals) is due to their ability to improve the efficiency of information sharing through collaborative filtering mechanisms for viral marketing recommendations and information dissemination techniques. Indeed, this information is disseminated from one node to another in a self-replicating manner, by sharing it with friends on the social network [9]- [12]. Friendly relationships between social users can result in significant variation in social networks, as users are influenced by their friends in decision-making [13]. ...
With the emergence of social networks and their adoption by a large number of users, the importance of influencers continues to grow and companies are in a frantic race to recruit those most likely to promote their reputation and brand image. However, in the existing literature, there is little work that conducts quantitative studies on this subject in developing countries. For this reason, we conducted a study that attempts to understand the importance of influencers in reshaping public opinion of a company or brand. We chose as a subject of study a large Moroccan company operating in the telecommunications sector that hired a popular influencer among young Moroccans. We then adopted an approach based on scraping and analyzing the occurrences of the influencer's posts on Instagram and the content of the company's website and then publishing a questionnaire to 180 respondents in the age range of most of the followers of the influencer in question. The results suggest that a positive relationship exists between the influencer and brand reputation, meaning that if the person is following the influencer who has published content on the brand, that person is expected to be systematically aware of the brand, and vice versa.
... The massive adoption of social networks by users (especially professionals) is due to their ability to improve the efficiency of information sharing through collaborative filtering mechanisms for viral marketing recommendations and information dissemination techniques. Indeed, this information is disseminated from one node to another in a self-replicating manner, by sharing it with friends on the social network [9]- [12]. Friendly relationships between social users can result in significant variation in social networks, as users are influenced by their friends in decision-making [13]. ...
With the emergence of social networks and their adoption by a large number of users, the importance of influencers continues to grow and companies are in a frantic race to recruit those most likely to promote their reputation and brand image. However, in the existing literature, there is little work that conducts quantitative studies on this subject in developing countries. For this reason, we conducted a study that attempts to understand the importance of influencers in reshaping public opinion of a company or brand. We chose as a subject of study a large Moroccan company operating in the telecommunications sector that hired a popular influencer among young Moroccans. We then adopted an approach based on scraping and analyzing the occurrences of the influencer's posts on Instagram and the content of the company's website and then publishing a questionnaire to 180 respondents in the age range of most of the followers of the influencer in question. The results suggest that a positive relationship exists between the influencer and brand reputation, meaning that if the person is following the influencer who has published content on the brand, that person is expected to be systematically aware of the brand, and vice versa.
... Guo, Wang, and Leskovec (2011) also suggest that people are more likely to purchase a product from a seller when their friends have already purchased from that vendor. In this circumstance, identifying a user who significantly influences other users in the social network is critical because influencers supply the standing necessary to convey a message to other users (Kiss & Bichler, 2008). Wang and Yu (2017) examine the effect of social influence on purchase behavior by analyzing word-of-mouth (WOM) communications among users. ...
This study examines the social influence of users on e-book purchases within a social network drawing on the structural equivalence model. Structural equivalence holds that higher levels of social influence exist among socially equivalent people (Burt, 1987). Using structural equivalence, network users were classified as either equivalent or inequivalent. Given that data about social relationships among people are often limited, to assign users to groups, link estimation utilized product choices to calculate network measures. With that framework, purchasing behaviors were predicted using various algorithms. Consistent with structural equivalence, the findings demonstrate that the average accuracy under the various algorithms is significantly higher in equivalent than inequivalent networks. Finally, comparing results with and without the network measurement variables suggests that failing to consider social equivalence may mislead prediction results by overestimating the effect of social influence of low equivalent groups or underestimating the effect of high social equivalent groups.
... We also checked the participation of celebrities, who are influential users that can affect network spread and activity (González-Bailón et al., 2013;Kiss & Bichler, 2008). To identify such influential users in our dataset we have used a database of 71 706 celebrities (individuals with a verified Twitter account and a designated Wikipedia article) compiled by Wiegmann et al. (2019). ...
This article explores how successful digitally native activism generates social change. Digitally native movements are initiated, organized, and coordinated online without any physical presence or pre-existing offline campaign. To do so, we explore the revelatory case of Sleeping Giants (SG)—an online movement that led more than 4,000 organizations to withdraw their programmatic advertising spend from Breitbart, a far-right publisher. Analyzing 3.5 million tweets related to the movement along with qualitative secondary data, we used a mixed method approach to investigate the conditions that favored SG emergence, the organizing and coordinating practices of the movement, and the strategic framing practices involved in the tuning of the movement’s language and rhetoric toward its targets. Overall, we contribute to research on online movements and shed light on the pivotal role of peer production work and of language in leading an impactful online movement that aimed to counter online disinformation and hate speech.
... Cook and Sheeran (2004) identify subject matter experts, journalists and other semi-public figures, and highly visible public figures, as being amongst the most notable influencers. In turn, different influencers are believed to impact upon many different people in many different ways (Kiss and Bichler, 2008). Studies indicate that such influencers are content creators, characterized by their posting of blogs, videos and so forth (Booth and Matic, 2011). ...
This study conducts an analysis of social media discussions related to high engagement sports brands. More specifically , our study examined the English Premier League (EPL); it sought to retrieve data systematically over the same day, weekly, for a period of five months. After this process, we had built 20 datasets and NodeXL was utilized to analyse the data. After we had this data, we were able to use qualitative observations to identify key users and conversations that formed around the EPL as well as the connections between the conversations that arose from the brand's posts and the people involved in them. We also analyzed the quantitative data underpinning our network visualisations to provide further insights. The most obvious initial finding was that when the EPL tweets, it prompts a large volume of conversations directly related to these tweets. However, we also noted that EPL tweets also help instigate further, sometimes unrelated, tweets and conversations. More specifically, we identified that the visualized network of conversations was of a broadcast form, which is characterized by messages being generated by a central account (the EPL) and shared by a number of decentralized users. Based on our analysis, we propose guidance around (S)ocial media presence, (C)rafting the message, Planned (i)ntervention, (S)pontaneous follow-up, and (M)essage mortality to form the SCISM framework. This framework is likely to be of interest to brands that wish to promote, sustain and benefit from their instigation of social media.
... The PageRank algorithm, for example, which is the fundamental search engine mechanism of Google, uses the topology of the web as an indicator of the value attached to any page (Brin & Page, 1998). Using a number of computational experiments on artificial and real networks in call data from a telecom company, Kiss and Bichler (2008) observed a significant increase in message diffusion when using influencers. Several studies have measured social influence by counting how much information related to a topic can be diffused in a network. ...
... Literature indicates there is more trust in influencers than other figures (Kiss & Bichler, 2008), since they are, in many ways, similar to their audience (Uzonoglu & Kip, 2014), regarded as authentic (Petrescu et al., 2018) responses and found that verification did not influence evaluations of source credibility. ...
Media giants, among them Facebook, Instagram and Twitter, support verified accounts. Verification, denoted by a blue checkmark badge visible in search and on one’s profile, is ostensibly a way of confirming one’s identity, yet only accounts with large followings are awarded verification status by the platform. This research investigates the perception of verification in the context of paid partnerships with social media influencers, a topic relatively absent from the literature despite the billions of dollars spent on influencer partnerships. Verified influencers cost more, therefore, this research could allow brands to capitalize their ad return if they are made aware of the implications associated with verification. Specifically, I investigate if consumers perceive verification as more directly associated with credibility or celebrity and if this relationship yields discrepancies in consumer’s trust of the brand, advertisement, and endorser in paid partnerships on social media. Two questionnaires administered via Amazon’s Mechanical Turk tested two hypotheses. 342 respondents completed a pre-test that tested, and proved true, the assumption that verification is viewed as the same regardless of platform. In the primary study, 413 participants were randomly assigned to one artificial Instagram post in a 2 x 2 between-subject design: (beauty vs. fitness industry) x (verified vs. unverified). Surprisingly, results indicated that verification had no impact on user’s perceptions of credibility, celebrity or trust. Interestingly, verification did play a significant role in user’s perceptions of endorser attractiveness and beauty and verified endorsers were viewed as less attractive. Given the findings, supplemental, future research is discussed as well as implications for marketers since verified endorsers showed no statistically significant benefits, yet they are costlier to work with.
... Consumers' Social Network. Consumers' social networks in the real world are different from those formed by Internet users [56]. Many empirical analyses show that social networks in the real world are mainly represented by smallworld networks, and most of the social networks in the internet world are scale-free networks [40,57]. ...
In recent years, green product issues have received increasing attention. Both government regulations and consumer behaviors have a strong influence on the product green degree decisions of manufacturers’ products. In order to find how government regulations and individual’s green product purchase behavior affect manufacturers’ green degree decisions and the market evolution characteristics, this paper proposes a multiagent model that considers the interaction among government, consumers, and manufacturers. The simulation results show that, firstly, the product green degree decision-making of manufacturers needs the guidance and regulation of the government. Secondly, product price subsidies are the most effective way to affect the manufacturers’ product green degree decisions. In contrast to giving green cost subsidies to manufacturers, the government employs various publicity means to improve the environmental awareness of consumers is also an effective way to enhance the green degree of manufacturers’ products. Thirdly, there is a “Crowding Out Effect” on the other qualities of manufacturers’ products when manufacturers focus on the green degree of their products.
... Given the predominance of these services in our daily lives and their universal diffusion (especially within the young generations [19]), the large impact on social media marketing strategies, and the spreading potential of these highly influential nodes [20], it is important to understand (i) how the UGC relates with the emergence of tremendously fast-growing social media influencers, and (ii) what are the properties of the resulting networks. ...
Many of today’s most used online social networks such as Instagram, Youtube, Twitter, or Twitch are based on User-Generated Content (UGC), and the exploration of this content is enhanced by the integrated search engines. Prior multidisciplinary effort on studying social network formation processes has privileged topological elements or socio-strategic incentives. Here, we propose an untouched meritocratic approach inspired by empirical evidence on Twitter data: actors continuously search for the best UGC provider. We statistically and numerically analyze the network equilibria properties: while the expected outdegree of the nodes remains bounded by the logarithm of the network size, the expected indegree follows a Zipf’s law with respect to the quality ranking. Notably, our quality-based mechanism provides an intuitive explanation of the origin of the Zipf’s regularity in growing social networks. Our theoretical results are empirically validated against large data-sets collected from Twitch, a fast-growing platform for online gamers.
... The PageRank algorithm, for example, which is the fundamental search engine mechanism of Google, uses the topology of the web as an indicator of the value attached to any page (Brin & Page, 1998). Using a number of computational experiments on artificial and real networks in call data from a telecom company, Kiss and Bichler (2008) observed a significant increase in message diffusion when using influencers. Several studies have measured social influence by counting how much information related to a topic can be diffused in a network. ...
An increasing amount of research on Intelligent Systems/Artificial Intelligence (AI) in marketing has shown that AI is capable of mimicking humans and performing activities in an ‘intelligent’ manner. Considering the growing interest in AI among marketing researchers and practitioners, this review seeks to provide an overview of the trajectory of marketing and AI research fields. Building upon the review of 164 articles published in Web of Science and Scopus indexed journals, this article develops a context-specific research agenda. Our study of selected articles by means of Multiple Correspondence Analysis (MCA) procedure outlines several research avenues related to the adoption, use, and acceptance of AI technology in marketing, the role of data protection and ethics, the role of institutional support for marketing AI, as well as the revolution of the labor market and marketers’ competencies.
50 days' free access - no sign-up, registration, or fees are required – available at the following link: https://www.sciencedirect.com/science/article/pii/S0148296321000643?dgcid=author
... This ability is usually multiplied through social networks or other online platforms (Anderson 2017;Hayta 2013;Valenzuela 2013). People that possess the power to exert digital influence have been coined influencers (Kiss and Martin 2008;Uzunoglu and Misci Kip 2014). Though this term most commonly refers to people with a large group of followers who receive payment for promoting a product, influencers may also be real-life friends or online reviewers who lead a consumer to purchase a certain product through positive reviews (Jiménez and Mendoza 2013). ...
Digitalization is leading to profound changes in our private and work lives. New technologies are pervasive and create opportunities for new business models and lifestyles. Recently, the term “Corporate Digital Responsibility” has been coined to summarize the emerging responsibilities of corporations relating to their digitalization-related impacts, risks, challenges, and opportunities. The paper at hand reviews the topic of CDR using a multi-step approach. First, results from an opinion poll of 509 US-based respondents are reported which illustrate the perceived opportunities and threats associated with the topic of digitalization, underlining the need for a strategic approach to CDR implementation. Second, existing uses and definitions of the CDR terminology are summarized and a definition of CDR is derived. Third, twenty important topics related to CDR are identified, summarized and categorized into three categories using the ESG (Environmental, Social, Governance) framework. Finally, results are discussed with regards to their theoretical and managerial contributions and a hands-on guide which companies can use to implement a suitable CDR strategy is presented.
... Kiss and Bichler [47] compared the performances of seven existing centrality measures including SenderRank, a new technique which was developed by the authors, to identify influential users in a social network constructed from a dataset collected from a telecom company. SenderRank and Out-degree centrality performed well in determining the most central nodes in a network of calls from a telecommunication company. ...
The number of mobile phone users is increasing tremendously. The social interaction between these mobile phone users can be represented using social network graphs. This type of study has very important applications in various areas especially in the detection of criminal groups who also use these devices to interact and plan their activities. Moreover, the study of identifying influential nodes in social network of any kind is currently receiving attention in the research arena. This is because identification of influential nodes of any network is significant to understanding the network. This becomes very important if the network in question is a criminal network, considering the insecurities of the current time. In this paper, a survey of influential nodes detection methods is carried out, we first define the problems associated with influential nodes detection and then examine various methods of identifying influential nodes. We also consider techniques employed in analysing users in the mobile phone network.
... The most usual metrics in the literature and in real-life problems belong to two classes: centralized and link topological metrics (see, e.g., [Kiss andBichler 2008, Peng et al. 2018]). Centralized metrics characterize the spread capabilities of the nodes and also describes nodes' proximity to the other players in the network; while link topological metrics emphasize important neighbors, benefiting from their relevancy. ...
Influence propagation in social networks is a subject of growing interest. A relevant issue in those networks involves the identification of key influencers. These players have an important role on viral marketing strategies and message propagation, including political propaganda and fake news. In effect, an important way to fight malicious usage on social networks is to understand their properties, their structure and the way messages propagate. This paper proposes two new indices for analysing message propagation in social networks, based on the network topological nature and the power of the message. The first index involves the strength of each node as a launcher of the message, dividing the nodes into launchers and non-launchers. The second index addresses the potential of each member as a receiver (target) of the message, dividing the nodes into targets and non-targets. Launcher individuals should indicate strong influencers and target individuals should identify the best target consumers. These indices can assist other known metrics when used to select efficient influencers in a social network. For instance, instead of choosing a strong and probably expensive member according to its degree in the network (number of followers), we may previously select those belonging to the launchers group and look for the lowest degree members, which are probably cheaper but still guarantying almost the same influence effectiveness as the largest degree members. On a different direction, using the second index, the strong target members should characterize relevant consumers of information in the network, which may include fake news' regular collectors. We discuss these indices using small-world randomly generated graphs and a number of real-world social networks available in known datasets repositories.
Companies communicate their brands with customers through social media, either officially through their official pages or non-officially through social media users’ Electronic Word of Mouth or brand-related content created by Social Media Influencers. This study evaluates the effectiveness of these three brand communication methods on consumer-based brand equity (CBBE), including brand awareness, brand image, brand attitude and purchase intention. The study further clarifies the most influential consumer and firm-related factors on Egyptian women’s behavioural intentions. It identifies the motives for following the official and non-official brand communication methods, interaction with brand posts and characteristics of the most followed social media influencers. A total of 400 Egyptian women answered an online/offline survey. Moreover, four focus group discussions were conducted. Based on the modified brand value chain model, the findings indicate that users’ eWOM is the most followed and influential brand communication method in Egyptian women’s purchase intention. Most Egyptian women are silent followers and prefer to follow younger influencers interested in fashion, sports, travelling and visiting new places. Egyptian women’s age, working status, past brand experience, real need to purchase, visuals of brand posts and peer pressure are the most influential firm-related and consumer-related variables on purchase intention.
It is becoming more and more promising that marketers hire influencers to launch campaigns for spreading items (e.g., articles or videos about products) over social media platforms. Such social media influencer marketing may generate tremendous utility if the influencers persuade their followers to adopt the recommended items. This could further spur extensive spontaneous item propagation on social media. Although prior studies mainly focus on influencer-selection strategies by the influencers’ traits, marketers with a number of items are often requested to determine both influencers and marketing items. The appropriateness between influencers and items is critical, but rarely considered in prior influencer-identification methods. We thus formulate and solve a novel cost-effective social media influencer marketing problem to maximize marketers’ utility by selecting appropriate pairwise combinations of influencers and items (i.e., item-influencer pairs). In particular, we first model utility functions and propose a simulation-based method to estimate the appropriateness of arbitrarily given item-influencer pairs by their potential utility. With the estimated utility, we devise an algorithm to iteratively select appropriate item-influencer pairs under various realistic conditions, including marketers’ budget, influencers’ payments, item-user fitness, social propagation, and influencers’ marketing slots. We theoretically prove that the marketing utility achieved by our method is near-optimal. We also conduct extensive empirical experiments with three real-world data sets to verify the superiority of our method in terms of cost-effectiveness and computational efficiency. Lastly, we discuss insightful theoretical and practical implications.
History: Accepted by Ram Ramesh, Area Editor for Data Science and Machine Learning.
Funding: This study was partially funded by the National Natural Science Foundation of China [Grants 72071125, 72031001, and 61972008].
Supplemental Material: The online appendix is available at https://doi.org/10.1287/ijoc.2022.1246 .
Cette recherche a pour objectif de cerner les déterminants de partage de Brand Content Digital (BCD) par les e-Leaders d'opinion sur les réseaux socioprofessionnels. Ce choix repose sur deux constats ; premièrement, la littérature sur les Médias Sociaux, s'est centrée sur la catégorie des Réseaux Socionumériques RSn (e.g. Facebook, Instagram, Myspace, Friendster…) au détriment de la catégorie des sites de réseautage professionnel (LinkedIn, Viadeo, Xing, SkilledAfricans…), malgré leur importance dans les contextes B2B et B2G. Deuxièmement, le partage de Brand Content Digital (BCD) est effectué par une infime partie des membres de ces réseaux socioprofessionnels, qui sont des Leaders d'opinion digitaux, dont les motivations semblent différer de ce que propose la littérature à propos des membres ordinaires de ce type de réseaux. Dans une perspective hypothético-déductive, la recherche s'est déployée en trois phases. Une première phase qualitative a permis de conduire six entretiens en profondeur avec les experts et les gestionnaires des réseaux socioprofessionnels en France (i.e. LinkedIn et Viadeo), complétés par 5 mini-focus groupes menés dans deux villes différentes (Angoulême et Poitiers). Cette première phase d'analyse qualitative a permis d'élaborer un modèle conceptuel des antécédents de partage de BCD par les e-Leaders d'opinion, fondé sur la Théorie du Comportement Planifié (TCP). Ensuite, ce modèle a été testé à travers 312 questionnaires en ligne administrés auprès de e-Leaders d'opinion, sélectionnés selon une nouvelle approche méthodologique, dite métrique-algorithmique. Les résultats de cette recherche permettent d'apporter différents types de contribution au débat scientifique. Tout d'abord, elle conduit à valider la structure de la TCP et sa force explicative sur la catégorie des réseaux socioprofessionnels, en soulignant que l'e-réputation et le gain social sont les facteurs attitudinaux clés expliquant la propension des e-Leaders d'opinion à s'engager dans le partage de BCD. En outre, ce travail permet de déterminer certaines motivations contextuelles propres tant aux Leaders d'opinion digitaux qu'à la catégorie des réseaux socioprofessionnels. Cette recherche offre également un certain éclairage sur le fonctionnement et l'évolution du rôle modérateur des traits de personnalité des leaders d'opinion digitaux sur l'intention et l'expérience de partage de BCD.A la lumière des résultats obtenus, plusieurs pistes et implications managériales sont proposées, permettant ainsi d'assoir et assister les webmanagers des entreprises dans l'optimisation de la diffusion de leurs campagnes BCD, en déterminant notamment sur quelle variable et dans quelle mesure il faut agir, observer et contrôler.
Purpose
There are two major strategies for short video advertising which are KOL (key opinion leader) endorsement and in-feed advertising. The authors aim to research the effectiveness of these two strategies for heterogeneous sellers.
Design/methodology/approach
The study employed a data set of users from Douyin. Using an endogenous treatment model, the study empirically examines the two strategies' effectiveness in attracting product traffic for online retailors at a short video app Douyin (TikTok).
Findings
The results show that the performance of in-feed advertising is higher when the seller's product is of lower price and when the seller has smaller cumulative video exposure. In addition, KOL endorsement is effective regardless of the product price, but performs better when the seller has larger cumulative video exposure.
Originality/value
To the best of the authors’ knowledge, this study is one of the first to explore the interaction effects of two major advertising strategies, KOL endorsement and in-feed advertising on short video platforms. The findings provide important theoretical contributions and practical implications.
Incentives for adopters and senders increase rates of product adoption and message forwarding respectively. In this paper, a dual incentives strategy is investigated to determine the optimized incentive combination in the new customer acquisition stage for minimizing cost and the optimal incentive combination in the repeated purchase stage for maximizing long-term profit. We construct a two-stage N-intertwined model, i.e., the SANB model to describe these two processes. Because of the advantage of the N-intertwined model in expressing the individual-level state transition process, heterogeneous incentives are also investigated. Results of homogeneous incentives show a more important role of promotion incentives. While results of heterogeneous incentives show different peculiarities between adoption and promotion incentives without the constraint of the total cost. Besides, a dataset from an online food delivery service platform is applied in an empirical study to verify the influence of incentives on exposure and purchase.
In this paper, we investigate whether transportation network centrality determines housing price in cities. We find that it does. Using housing price data from >400 neighborhoods across the city of Kolkata, India, our research shows that a neighborhood's centrality within the intra-urban road transportation network is positively associated with the average price per sq. ft. of ownership units in multistory apartment buildings in the neighborhood. We test the relationships of three alternative centrality indices – closeness, betweenness, and eigenvector – capturing different dimensions of network-wide connectedness with housing price, independently and in combination. We employ two alternative network weights to derive centrality considering peak-period and off-peak travel conditions and road transportation network performance. We address the spatial autocorrelation issue to derive robust evidence on the centrality-price relationship. Our results suggest that centrality is an intrinsic location advantage that positively influences urban housing price. Different centrality indices have different effects on price but they collectively reinforce each other. The estimated magnitudes of association between centrality indices and housing price are significant for policy and practice. In addition to contributing to scholarship in the domain of transportation planning, our research offers specific takeaways for metropolitan planning agencies and real-estate developers, especially in resource constrained geographic contexts.
Social media platforms are the key tools to facilitate online engagement; however, to stimulate a discussion, the content published on the platforms is significant as it must appeal to different consumers. The quality of the content and platform type is key to successful engagement. Maintaining positive relationships with consumers is a vital activity for many brands in social media. Trust, satisfaction, fairness, and mutual dependency are key factors to retaining customers. Moreover, positive brand attitudes and higher purchase intentions were found to be linked to positive evaluations of companies' social media postings. To maintain value, firms use social media platforms that facilitate consumer-to-consumer as well as consumer-to-business engagement. Drawing from social influence theory, this chapter explores how social media marketing content (SMMC) impacts customer retention.
As social media usage has grown over the recent past, so too has a new form of celebrity: social media influencers (SMIs). It can be difficult for consumers to know whether the influencers they are following are real. To combat this, social media sites introduced user verification. Verification, denoted by a blue checkmark badge visible in search and on one's profile, is ostensibly a way of confirming one's identity, yet only accounts with large followings are awarded verification status by the platform. In this paper, we investigate the perception of verification in the context of SMIs, a topic relatively absent from the literature despite the billions of dollars spent on partnerships. Specifically, we investigate if consumers perceive verification as more directly associated with credibility or celebrity. Further, we consider whether the fit of the influencer with the advertisement yields discrepancies in consumers' trust of the advertisement and endorser in paid partnerships on social media. Through two studies, we find that consumers associate verification most closely with celebrity, rather than authenticity and that when the influencer is advertising a product that does not fit with their brand, consumers are significantly less likely to trust verified accounts rather than unverified accounts. This study has theoretical implications for marketing researchers as well as practical implications for marketing managers. Verified influencers cost more; therefore, this research provides unique insights for brands to capitalize their ad return if they are made aware of the implications associated with verification.
Seeding is the introduction of a new product to a portion of consumers (seeds) before it is widely introduced to the market, and is one of the most important drivers in the design of marketing campaigns and effective word-of-mouth advertising. This research aims to identify different dimensions of seeding through meta-synthesis. For this purpose, 68 articles that had studied the subject of seeding were selected and coded on the concepts extracted from them. In the analysis of chosen researches, from 189 primary codes, 90 indicators (selective code), 18 components (sub-dimension), and five dimensions (main component) were identified. For the first time in this paper, different dimensions of seeding were investigated by the meta-synthesis method. According to the reviewed articles, the concepts extracted from the literature are: 1) Factors affecting seeding with 33 indicators and seven components that are target community characteristics, target community homogeneity, message nature, environmental conditions, internal conditions, marketing tools, and implementation strategies; 2) Characteristics of seeding target market with 21 indicators and three components that are consumer characteristics, customer nature and characteristics of individuals; 3) Communication and networking of seeding with 15 indicators and three components that are a big picture of seeding, position of seeding and nature of social network; 4) Seeding and growth and development with 12 indicators and three components that are value creation, market growth and development and product growth and development; And finally 5) Seeding and marketing with nine indicators and two components that are marketing tools strengthen and customer experience strengthen.
Presently, people use social media at a greater rate to share their personal investment experiences. This plentiful user-generated data source has been promisingly used by investors for portfolio creation. A new type of investing platform that allows investors to copy the portfolios of experienced investors has grown dramatically. In this research, we propose a collective intelligence mechanism that can extract and consolidate the opinions expressed over the social investing platform and generate appropriate portfolios by analyzing other investors’ knowledge, authority, and opinions toward the investment target. The experimental results obtained based on the social investing platform eToro.com reveal that the portfolio recommended by the proposed mechanism outperforms the market index and other benchmark approaches in various financial performance aspects.
Amid the skyrocketing increase in social media content, understanding the information-sharing mechanism is crucial. We analyze Twitter messages on recent public issues and IT trends to examine the network's bi-directional message diffusion patterns through incoming and outgoing ties, based on the users’ retweeting and mentioning behavior toward one another. We find context-based patterns. This study shows that regarding IT trends, the use of the “like” feature significantly related to users’ influence, where outgoing ties had a greater mediating influence than incoming ties. During COVID-19, users prioritized obtaining reliable information. Our findings contribute toward developing diffusion mechanisms and social media strategies.
Since the past five years, the size of social networks as well as social media users have been doubled and this growth is creating a new marketing hurdle i.e. noise. Suddenly, the paid advertisements on social media have become less effective because of the noise that has been developed by these online users. This is the reason that 615 million social media users are using ad-blocking software and applications which is growing by 30% every year. On one hand, users want to have an organic and genuine connection with the brand, but this noise is not allowing them to do so. While on the other hand marketers are reconsidering their advertising strategies and finding new and more effective techniques of marketing. It is the reason that marketers are striving hard to find the best way that would influence consumers as well as bring sales conversion. The best way to do this is to focus on consumer-oriented marketing strategies such as Product placement, native advertising, content marketing, and influencer marketing to avoid the noise that is created on social media. Social media influencer marketing is a term that refers to leveraging the ability of key people to support a brand and spread the word to their followers. It has been established as a new as well as a highly effective method for brands to build and engage with audiences on social media. This paper focuses on the evolution of influencer marketing from traditional marketing i.e. e-WOM from WOM derived from various kinds of literature available on the internet.
Influencer marketing has become a powerful channel for brand promotion and market expansion in the hospitality industry. However, those responsible for implementing influencer marketing campaigns are susceptible to the “myth of viewability” and rely on the Cost Per View (CPV) evaluation metric, rather than the more appropriate Cost Per Action (CPA). The current research explores the aforementioned myth from a hospitality management perspective by identifying the types of image (or photo) which attract more audience commentary or liking. A qualitative research approach is adopted involving two experiments with influencer pairs across the restaurant context in Taipei, Taiwan. We selected influencers Q and S as our manipulated group. In sharing images on their social media platform, it was found that they make greater use of personal than of food related images. The opposite was the case for the controlled group - influencers X and A – who shared more food than personal images. The researchers tracked viewer responses and then actions towards influencer postings to determine (a) which influencer approach draws more views and (b) the costs that are attributable to views and/or actions. They drew upon the findings to formulate an Owner-Influencer Matrix, a strategic planning tool and framework that helps owners and influencers to optimize influencer marketing. It is concluded that interactions between influencers and business owners should be beneficial to both parties. This empirical study may provide business owners and social media influencers with insights about communicating the respective brand values of their counterparts and designing sponsorship collaborations with a capacity to generate the desired consumer responses.
Shareholder valuations are economically and statistically positively correlated with independent director power, gauged by a composite of social network power centrality measures. Powerful independent directors’ sudden deaths reduce shareholder value significantly; other independent directors’ deaths do not, consistent with powerful independent directors increasing firm valuations. Further tests associate more powerful independent directors with less value-destroying M&A, less free cash flow retention, more CEO accountability, and less earnings management. We interpret these findings as more powerful independent directors better detecting and countering CEO missteps because of better access to information, greater credibility in challenging errant top managers, or both.
This article is protected by copyright. All rights reserved
Understanding the growth paths of artificial intelligence (AI) and its impact on branding is extremely pertinent of technology-driven marketing. This explorative research covers a complete bibliometric analysis of the impact of AI on branding. The sample for this research included all 117 articles from the period of 1982-2019 in the Scopus database. A bibliometric study was conducted using co-occurrence, citation analysis and co-citation analysis. The empirical analysis investigates the value propositions of AI on branding. The study revealed the nine clusters of co-occurrence: Social Media Analytics and Brand Equity; Neural Networks and Brand Choice; Chat Bots-Brand Intimacy; Twitter, Facebook, Instagram-Luxury Brands; Interactive Agent-Brand Love and User Choice; Algorithm Recommendations and E-Brand Experience; User-Generated Content-Brand Sustainability; Brand Intelligence Analytics; and Digital Innovations and Brand Excellence. The findings also identify four clusters of citation analysis—Social Media Analysis and Brand Photos, Network Analysis and E-Commerce, Hybrid Simulating Modelling, and Real-time Knowledge-Based Systems—and four clusters of co-citation analysis: B2B Technology Brands, AI Fostered E-Brands, Information Cascades and Online Brand Ratings, and Voice Assistants-Brand Eureka Moments. Overall, the study presents the patterns of convergence and divergence of themes, narrowing to the specific topic, and multidisciplinary engagement in research, thus offering the recent insights in the field of AI on branding.
Despite its increasing role in communication, the world wide web remains the least controlled medium: any individual or institution can create websites with unrestricted number of documents and links. While great efforts are made to map and characterize the Internet's infrastructure, little is known about the topology of the web. Here we take a first step to fill this gap: we use local connectivity measurements to construct a topological model of the world wide web, allowing us to explore and characterize its large scale properties. Comment: 5 pages, 1 figure, updated with most recent results on the size of the www
We present a study of information flow that takes into account the observation that an item relevant to one person is more likely to be of interest to individuals in the same social circle than those outside of it. This is due to the fact that the similarity of node attributes in social networks decreases as a function of the graph distance. An epidemic model on a scale-free network with this property has a finite threshold, implying that the spread of information is limited. We tested our predictions by measuring the spread of messages in an organization and also by numerical experiments that take into consideration the organizational distance among individuals.
The author explores the information needs of service consumers. In the purchase decision process, search behavior is motivated in part by perceived risk and the consumer's ability to acquire relevant information with which purchase uncertainty can be addressed. Marketing theory suggests that consumers use information sources in a distinctive way to reduce the uncertainty associated with services. Hence, six hypotheses are developed to test the information acquisition of service buyers. An experimental approach is employed to compare, in a prepurchase context, the information sources used by consumers of services and those used by consumers of goods. The resulting data support the predictions offered and extend marketing theory.
This paper presents two measures, integration and radiality, which indicate the degree an individual is connected and reachable within a network. The measures are created using a reverse distance (geodesic) matrix, thus providing a directed closeness measure. Integration and radiality are then correlated with other centrality measures to examine the differences and similarities between each other and these other centrality measures. We then show how integration and radiality are associated with certain demographic and substantive variables. Integration and radiality provide a new index of network structure and our results show how that structure may influence behavior.
The ability to measure centrality in social networks has been a particularly useful development in social network analysis. For researchers trying to decide which centrality measure is most meaningful and valid for their research purposes, various papers have explored the conceptual foundations of centrality measures. Less well documented is the empirical performance of centrality measures under different research scenarios or constraints. This study uses bootstrap sampling procedures to determine how sampling affects the stability of 11 different network centrality measures. Results indicate that some measures are more stable than others, and that stability is also a function of network and study properties.
Centrality measures, or at least popular interpretations of these measures, make implicit assumptions about the manner in which traffic flows through a network. For example, some measures count only geodesic paths, apparently assuming that whatever flows through the network only moves along the shortest possible paths. This paper lays out a typology of network flows based on two dimensions of variation, namely the kinds of trajectories that traffic may follow (geodesics, paths, trails, or walks) and the method of spread (broadcast, serial replication, or transfer). Measures of centrality are then matched to the kinds of flows that they are appropriate for. Simulations are used to examine the relationship between type of flow and the differential importance of nodes with respect to key measurements such as speed of reception of traffic and frequency of receiving traffic. It is shown that the off-the-shelf formulas for centrality measures are fully applicable only for the specific flow processes they are designed for, and that when they are applied to other flow processes they get the “wrong” answer. It is noted that the most commonly used centrality measures are not appropriate for most of the flows we are routinely interested in. A key claim made in this paper is that centrality measures can be regarded as generating expected values for certain kinds of node outcomes (such as speed and frequency of reception) given implicit models of how traffic flows, and that this provides a new and useful way of thinking about centrality.
We consider prediction-model evaluation in the context of marketing-campaign planning. In order to evaluate and compare models with specific campaign objectives in mind, we need to concentrate our attention on the appropriate evaluation-criteria. These should portray the model's ability to score accurately and to identify the relevant target population. In this paper we discuss some applicable model-evaluation and selection criteria, their relevance for campaign planning, their robustness under changing population distributions, and their employment when constructing confidence intervals. We illustrate our results with a case study based on our experience from several projects.
The analysis of biological networks involves the evaluation of the vertices within the connection structure of the network. To support this analysis we discuss five centrality measures and demonstrate their applicability on two example networks, a protein-protein-interaction network and a transcriptional regulation network. We show that all five centrality measures result in different valuations of the vertices and that for the analysis of biological networks all five measures are of interest.
The study of network topologies provides interesting insights into the way in which the principles on which interconnected systems are constructed influence the dynamics of diffusion and communication processes in many kinds of socio-technical systems. Empirical research has shown that there are principles of construction similar to those of the laws of nature for social networks and their technical derivatives, like E-mail networks, the internet, publication co-authoring, or business collaboration. For decades, the paradigm of a randomly connected network has been used as a model for real world networks, in ignorance of the fact that they are only a poor fit for such networks. Apparently, all the above-mentioned networks share the same building blocks. They attach new members over time and the attachment prefers existing members that are already well connected. This principle of “preferential attachment” leads to interesting properties that have to be taken into consideration when analyzing and designing systems with some kind of network background.
What are called “scale-free” networks seems to be a better fit for the description of real world networks. They use preferential attachment as a construction principle to resample real world networks. Their behavior in terms of diffusion and communication processes is fundamentally different from that of random networks.
To illustrate the potential value of the discovery of scale-free networks for applications in information systems related research, an example will be used in this article to illustrate their usefulness for realistic network modeling. A scale-free communication network of security traders will show what impact network topology has on the dynamics of complex socio-technical systems.
A Family of new measures of point and graph centrality based on early intuitions of Bavelas (1948) is introduced. These measures define centrality in terms of the degree to which a point falls on the shortest path between others and therefore has a potential for control of communication. They may be used to index centrality in any large or small network of symmetrical relations, whether connected or unconnected.
Systems as diverse as genetic networks or the World Wide Web are best described as networks with complex topology. A common property of many large networks is that the vertex connectivities follow a scale-free power-law distribution. This feature was found to be a consequence of two generic mech-anisms: (i) networks expand continuously by the addition of new vertices, and (ii) new vertices attach preferentially to sites that are already well connected. A model based on these two ingredients reproduces the observed stationary scale-free distributions, which indicates that the development of large networks is governed by robust self-organizing phenomena that go beyond the particulars of the individual systems.
This research studies the optimal seeding of a network of consumers who are potential adopters of a technology. This problem has a wide range of applications which include launching a network good, inducing network-based marketing, igniting collaboration in organizations, and stimulating R&D with local spillovers. I have two main results so far. The first result shows that when a firm is unaware of how connected its consumers' are, randomly seeding a network is almost always optimal. The second result provides a a simple and plausible set of conditions under which, when a firm that is aware of how connected its consumers are, it is optimal for the firm to "seed the fringes rather than the hubs", that is, focus on seeding the least connected consumers, and not the highly connected consumers. These results hold for any distribution over any arbitrary set of asymmetric networks, and are also independent of the strength of the network effects. My ongoing research focuses primarily on extending these results to include seeding with probabilistic adoption, costly and constrained seeding, myopic consumers, and expanding the model to permit the social network to be a basis for the diffusion of product information. 2 Overview and outline of model When launching a network good, it often seems sensible for a seller to "seed" its network by giving away the product for free to a subset of its potential customers. If the good displays local network effects — that is, rather than being influenced by an increase in the size of a product's user base or network in general, if each consumer is influenced directly by the decisions of only a typically small subset of other consumers, for instance those he or she is "connected" to via an underlying social or business network — then the seller may wish to seed consumers based on how connected they are. Targeting more highly connected consumers seems like a sensible strategy, since these consumers are more influential — they influence and increase the willingness-to-pay of a larger subset of others. However, there is a trade-off: consumers who are more connected are also likely to have a higher eventual (post-seeding) willingness to pay for the network good, and are also more likely to be influenced by the seeding of others, on account of having a higher number of neighbors. 1 I thank seminar participants at Carnegie-Mellon University for their feedback.
Thisdissertation proposes a new epidemiological model to account forparticular characteristics of computer worm epidemics. This newmodel, termed the Progressive Susceptible-Infected-DetectedRemoved(PSIDR) epidemiological model, incorporates newaspects related to the availability of antivirus signatures, to theexistence of direct immunization, and to the presence of a curingphase. Various costs are incorporated in the model, which allow usto determine the best strategies to fight ...
Excellent service is the foundation for services marketing, contend Leonard Berry and A. Parasuraman in this companion volume to "Delivering Quality Service." Building on eight years of research, the authors develop a model for understanding the relationship between quality and marketing in services and offer dozens of practical insights into ways to improve services marketing. They argue that superior service cannot be manufactured in a factory, packaged, and delivered intact to customers. Though an innovative service concept may give a company an initial edge, superior quality is vital to sustaining success. Berry and Parasuraman show that inspired leadership, a customer-minded corporate culture, an excellent service-system design, and effective use of technology and information are crucial to superior service quality and services marketing. When a company's service is excellent, customers are more likely to perceive value in transactions, spread favorable word-of-mouth impressions, and respond positively to employee-cross-selling efforts. The authors point out that a service company that does relatively little pre-sales marketing but is truly dedicated to delivering excellent quality service will have greater marketing effectiveness, higher customer retention, and more sales to existing customers than a company that emphasizes pre-sale marketing but falls short during actual service delivery. The focus of any company, they insist, must be customer satisfaction through integration of service quality throughout the entire system. Filled with examples, stories, and insights from senior executives, Berry and Parasuraman's new framework for effective marketing servicescontains the key to high-performance services marketing.
The abstract for this document is available on CSA Illumina.To view the Abstract, click the Abstract button above the document title.
The authors examine behavioral outcomes following a customer-initiated contact (CIC) with a manufacturer and develop a framework to explain the impact of vendor performance during a CIC on a customer's share of category requirements with a focal brand and word-of-mouth incidence following contact. The authors propose customer characteristics and context-specific factors that may relate to differences in the key characteristics of the underlying source model of share of category requirements and word of mouth. The authors then assess the overall importance of the explanatory variables in the source model and simultaneously test for systematic differences related to CIC-specific factors using survey data from more than 1700 CICs that involve more than 60 brands. A key assumption in much prior research that has examined customer-firm interactions is that CIC-specific factors, if they are included at all, create an automatic regularity that must be controlled for. The authors propose and find an additional effect. The responsiveness to factors under a firm's control varies across CICs, and therefore firms that adapt their processing have an advantage. Rather than provide a uniform response to all CICs, the authors' results offer managers several guidelines on how to customize their responses to the various CIC types and how to improve the efficiency and effectiveness of their firms' CIC management efforts.
Three competing hypotheses about structural centrality are explored by means of a replication of the early MIT experiments on communication structure and group problem-solving. It is shown that although two of the three kinds of measures of centrality have a demonstrable effect on individual responses and group processes, the classic measure of centrality based on distance is unrelated to any experimental variable. A suggestion is made that the positive results provided by distance-based centrality in earlier experiments is an artifact of the particular structures chosen for experimentation.
The author examines consumer affective responses to product/consumption experiences and their relationship to selected aspects of postpurchase processes. In separate field studies of automobile owners and CATV subscribers, subjects reported the nature and frequency of emotional experiences in connection with product ownership and usage. Analysis confirms hypotheses about the existence of independent dimensions of positive and negative affect. Both dimensions of affective response are found directly related to the favorability of consumer satisfaction judgments, extent of seller-directed complaint behavior, and extent of word-of-mouth transmission.
Considers the literature relating to the financial benefits of service quality, discussing direct benefits in terms of retaining existing customers, and indirect benefits in terms of attracting new customers and inducing customers to increase their usage levels. Points out that virtually no empirical work has been done exploring the indirect benefits of service quality. Presents an empirical study investigating the nature of customer attraction effects and usage effects in a market for which customer retention is not yet an important factor. Finds that in the market studied, service quality impacts consumer attraction only through the mechanism of word of mouth, but that usage rates are driven by service quality but not by advertising. Discusses the managerial implications of these findings; also includes a discussion article which raises many important methodological issues and a rejoinder from the authors.
Word-of-mouth communication (WOM) is a dominant force in the marketplace for services. However, the current body of research provides little insight into the nature of WOM in the service marketplace. Reports the results of a content-analytic study that provides insight into WOM’s content and the catalysts by which it is stimulated. The goal was to capture a series of “grounded events” from which broader patterns could be discerned. These grounded events were actual incidents of WOM as described by the recipients of a communication. Three content categories and ten catalyst categories are identified. Implications for managers are addressed.
Although the concept of centrality has been well developed in the social networks literature, its empirical development has lagged somewhat. This paper moves a step in that direction by assessing the performance of four centrality models under a variety of known and controlled situations. It begins by examining the assumptions underlying each model, as well as its behavior in a community influence network. It then assesses the robustness and sensitivity of each model under conditions of random and systematic variation introduced into this network.
2In an influential paper, Freeman (1979) identified three aspects of centrality: betweenness, nearness, and degree. Perhaps because they are designed to apply to networks in which relations are binary valued (they exist or they do not), these types of centrality have not been used in interlocking directorate research, which has almost exclusively used formula (2) below to compute centrality. Conceptually, this measure, of which c(ot, 3) is a generalization, is closest to being a nearness measure when 3 is positive. In any case, there is no discrepancy between the measures for the four networks whose analysis forms the heart of this paper. The rank orderings by the
Due to the proliferation of information systems and technology, businesses increasingly have the capability to accumulate huge amounts of customer data in large databases. However, much of the useful marketing insights into customer characteristics and their purchase patterns are largely hidden and untapped. Current emphasis on customer relationship management makes the marketing function an ideal application area to greatly benefit from the use of data mining tools for decision support. A systematic methodology that uses data mining and knowledge management techniques is proposed to manage the marketing knowledge and support marketing decisions. This methodology can be the basis for enhancing customer relationship management. q 2001 Elsevier Science B.V. All rights reserved.
The study of the Web as a graph is not only fascinating in its own right, but also yields valuable insight into Web algorithms for crawling, searching and community discovery, and the sociological phenomena which characterize its evolution. We report on experiments on local and global properties of the Web graph using two AltaVista crawls each with over 200 million pages and 1.5 billion links. Our study indicates that the macroscopic structure of the Web is considerably more intricate than suggested by earlier experiments on a smaller scale.
(This article originally appeared in Management Science, January 1969, Volume 15, Number 5, pp. 215–227, published by The Institute of Management Sciences.)
A growth model for the timing of initial purchase of new products is developed and tested empirically against data for eleven consumer durables. The basic assumption of the model is that the timing of a consumer's initial purchase is related to the number of previous buyers. A behavioral rationale for the model is offered in terms of innovative and imitative behavior. The model yields good predictions of the sales peak and the timing of the peak when applied to historical data. A long-range forecast is developed for the sales of color television sets.
We model the diffusion of innovations in markets with two segments: who are more in touch with new developments and who affect another segment of whose own adoptions do not affect the influentials. This two-segment structure with asymmetric influence is consistent with several theories in sociology and diffusion research, as well as many “viral” or “network” marketing strategies. We have four main results. (1) Diffusion in a mixture of influentials and imitators can exhibit a dip or “chasm” between the early and later parts of the diffusion curve. (2) The proportion of adoptions stemming from influentials need not decrease monotonically, but may first decrease and then increase. (3) Erroneously specifying a mixed-influence model to a mixture process where influentials act independently from each other can generate systematic changes in the parameter values reported in earlier research. (4) Empirical analysis of 33 different data series indicates that the two-segment model fits better than the standard mixed-influence, the Gamma/Shifted Gompertz, and the Weibull-Gamma models, especially in cases where a two-segment structure is likely to exist. Also, the two-segment model fits about as well as the Karmeshu-Goswami mixed-influence model, in which the coefficients of innovation and imitation vary across potential adopters in a continuous fashion.
In this article, the effect of word-of-mouth (WOM) communications on product judgments is investigated. Additionally, the moderating influence of several situational, personal, and source characteristics are studied in three experiments. These investigations show that WOM influences short-term and long-term judgments. This influence is greater when a consumer faces a disconfirmation experience and when the WOM communication is presented by an expert. Interestingly, personal characteristics such as susceptibility to interpersonal influence and product knowledge do not appear to moderate WOM.
Direct marketing firms want to transfer their message as efficiently as possible in order to obtain a profitable long-term relationship with individual customers. Much attention has been paid to address selection of existing customers and on identifying new profitable prospects. Less attention has been paid to the optimal frequency of the contacts with customers. We provide a decision support system that helps the direct mailer to determine mailing frequency for active customers. The system observes the mailing pattern of these customers in terms of the well-known R(ecency), F(requency) and M(onetary) variables. The underlying model is based on an optimization model for the frequency of direct mailings. The system provides the direct mailer with tools to define preferred response behavior and advises the direct mailer on the mailing strategy that will steer the customers towards this preferred response behavior.
This paper studies how a behavior spreads in a population. We consider a network of interacting agents whose actions are determined by the actions of their neighbors, according to a simple diffusion rule. We find, using a mean-field approach, the threshold for the spreading rate above which the behavior spreads and becomes persistent in the population. This threshold crucially depends on the connectivity distribution of the social network and on specific features of the diffusion rule.
Scale-free networks are abundant in nature and society, describing such diverse systems as the world wide web, the web of human sexual contacts, or the chemical network of a cell. All models used to generate a scale-free topology are stochastic, that is they create networks in which the nodes appear to be randomly connected to each other. Here we propose a simple model that generates scale-free networks in a deterministic fashion. We solve exactly the model, showing that the tail of the degree distribution follows a power law.
The co-authorship network of scientists represents a prototype of complex evolving networks. In addition, it offers one of the most extensive database to date on social networks. By mapping the electronic database containing all relevant journals in mathematics and neuro-science for an 8-year period (1991–98), we infer the dynamic and the structural mechanisms that govern the evolution and topology of this complex system. Three complementary approaches allow us to obtain a detailed characterization. First, empirical measurements allow us to uncover the topological measures that characterize the network at a given moment, as well as the time evolution of these quantities. The results indicate that the network is scale-free, and that the network evolution is governed by preferential attachment, affecting both internal and external links. However, in contrast with most model predictions the average degree increases in time, and the node separation decreases. Second, we propose a simple model that captures the network's time evolution. In some limits the model can be solved analytically, predicting a two-regime scaling in agreement with the measurements. Third, numerical simulations are used to uncover the behavior of quantities that could not be predicted analytically. The combined numerical and analytical results underline the important role internal links play in determining the observed scaling behavior and network topology. The results and methodologies developed in the context of the co-authorship network could be useful for a systematic study of other complex evolving networks as well, such as the world wide web, Internet, or other social networks.
The Web harbors a large number of communities — groups of content-creators sharing a common interest — each of which manifests itself as a set of interlinked Web pages. Newgroups and commercial Web directories together contain of the order of 20,000 such communities; our particular interest here is on emerging communities — those that have little or no representation in such fora. The subject of this paper is the systematic enumeration of over 100,000 such emerging communities from a Web crawl: we call our process trawling. We motivate a graph-theoretic approach to locating such communities, and describe the algorithms, and the algorithmic engineering necessary to find structures that subscribe to this notion, the challenges in handling such a huge data set, and the results of our experiment.
The intuitive background for measures of structural centrality in social networks is reviewed and existing measures are evaluated in terms of their consistency with intuitions and their interpretability.Three distinct intuitive conceptions of centrality are uncovered and existing measures are refined to embody these conceptions. Three measures are developed for each concept, one absolute and one relative measure of the centrality of positions in a network, and one reflecting the degree of centralization of the entire network. The implications of these measures for the experimental study of small groups is examined.
The Internet has become a rich and large repository of information about us as individuals. Anything from the links and text on a user’s homepage to the mailing lists the user subscribes to are reflections of social interactions a user has in the real world. In this paper we devise techniques and tools to mine this information in order to extract social networks and the exogenous factors underlying the networks’ structure. In an analysis of two data sets, from Stanford University and the Massachusetts Institute of Technology (MIT), we show that some factors are better indicators of social connections than others, and that these indicators vary between user populations. Our techniques provide potential applications in automatically inferring real world connections and discovering, labeling, and characterizing communities.
The concepts of the “center” and the “median vertex” of a graph are generalized to the “absolute center” and the “absolute median” of a weighted graph (a graph with weights attached to its vertices as well as to its branches). These results are used to find the optimum location of a “switching center” in a communication network and to locate the best place to build a “police station” in a highway system. It is shown that the optimum location of a switching center is always at a vertex of the communication network while the best location for the police station is not necessarily at an intersection. Procedures for finding these locations are given.
In assessing the potential of data mining based marlceting campaigns one needs to estimate the payoff of applying modeling to the problem of predicting behavior of some target population (e.g. attriters, people likely to buy product X, people likely to default on a loan, etc). This assessment has two components: a) the financial estimate of the campaign profitability, based on cost/benefit analysis and b) estimation of model accuracy in the targeted population using measures such as lift. We present a methodology for initial cost/benefit analysis and present surprising empirical results, based on actual business data from several domains, on achievable model accuracy. We conjecture that lift at T (where T is the target frequency) is usually about sqrt(llT ) for a good model. We also present formulae for estimating the entire lift curve and estimating expected profits. In assessing the potential of such campaigns one may want to estimate certain financial parameters such as profitability or relative increase in the concentration of the targets (lift) in the targeted group. In this paper we ask whether and when it is possible to quickly estimate parameters such as lift based on the problem features before attempting the complex task of modeling. We define guidelines for when one might consider deploying marketing campaigns and propose estimation formulae for lift at an important fixed point and over the entire lift curve. In sections 2 and 3 we define the basic terminology and guidelines for conducting such an assessment. In section 4 we investigate the relationship between lift and the target frequency. In section 5 we analyze how the lift decreases with respect to increasing subsets of population and in section 6 we use the lift curve and cost estimates to derive an estimate for expected maximum profit. We conclude with a discussion of the limits of these heuristics and possible directions for extending this inquiry.
y illustrative of the tremendous potential of KDD technology. 1.1 Risk Management and Targeted Marketing Insurance and direct-mail retail are examples of businesses that rely on effective data analysis in order to make profitable business decisions. For example, insurers must be able to accurately assess the risks posed by policyholders in order to set insurance premiums at competitive levels. Overcharging low-risk policyholders would motivate such policyholders to seek lower premiums elsewhere. Undercharging high-risk policyholders would attract more high-risk policyholders because of the lower premiums. In both cases, costs would increase and profits would decrease. Effective data analysis leading to the creation of accurate predictive models is essential in order to address these issues. In the case of direct-mail targeted marketing, retailers must be able to identify subsets of the population that are likely to respond to promotions in order to offset mailing and printing costs.
Part I. Introduction: Networks, Relations, and Structure: 1. Relations and networks in the social and behavioral sciences 2. Social network data: collection and application Part II. Mathematical Representations of Social Networks: 3. Notation 4. Graphs and matrixes Part III. Structural and Locational Properties: 5. Centrality, prestige, and related actor and group measures 6. Structural balance, clusterability, and transitivity 7. Cohesive subgroups 8. Affiliations, co-memberships, and overlapping subgroups Part IV. Roles and Positions: 9. Structural equivalence 10. Blockmodels 11. Relational algebras 12. Network positions and roles Part V. Dyadic and Triadic Methods: 13. Dyads 14. Triads Part VI. Statistical Dyadic Interaction Models: 15. Statistical analysis of single relational networks 16. Stochastic blockmodels and goodness-of-fit indices Part VII. Epilogue: 17. Future directions.