Network Effects Model Research Articles

Promoting the adoption of digital technology: Strategic policy insights from a network effects model

This paper examines a dynamic model of digital adoption by firms, incorporating network effects and highlighting the crucial role of government intervention to avoid a low stable equilibrium. The paper presents a strategic policy framework for promoting digital adoption effectively, offering three key insights for policy efforts targeting SMEs.First, there is a risk of ending up at a stable low equilibrium rate of digital adoption despite an initial surge in interest. Sustaining momentum beyond initial promotion efforts is essential. Second, it is critical to make a significant push when the digital adoption rate is approaching critical mass. This tipping point enables continuous, self-sustaining expansion. Third, understanding technology adoption cycles and allocating resources effectively is crucial. Once the adoption rate passes the critical threshold, further adoption occurs without promotional interventions. Resources should then be reallocated to foster the adoption of new technologies or applications, ensuring optimal use of limited resources to drive ongoing innovation and progress.The study also reveals that countries with a larger population of SMEs, due to either a large population or strong entrepreneurial environment, may find it relatively easier to promote digital adoption in this sector.

How Users Drive Value in Two-Sided Markets: Platform Designs That Matter

Extant research has popularized the perspective that strong network effects produce “winner-take-all” outcomes, which leads platforms to invest in user growth and encourages investors to subsidize these platforms. However, user growth does not necessarily imply strong user stickiness. Without user stickiness, strong network effects in the current period may fade in future periods, thus rendering a user growth strategy ineffective. By adding a time dimension to network effects, we developed a model of cross-period and within-period network effects to explain how different types of network effects drive value. We emphasize that the cross-period same-side network effect contributes to user stickiness, while the within-period cross-side network effect persists conditional on user stickiness. We propose that one reason for platforms having heterogeneous cross-period same-side network effects is because of the “product learning” mechanism: it is expected that products with higher uncertainty have a stronger cross-period same-side network effect. Based on different drivers, we extend the customer lifetime value model (CLV2) to two-sided platform markets, allowing us to measure how different interventions drive platform value. Using Groupon data, we verify our insights and discuss platform design choices that enhance user stickiness when the cross-period same-side network effect is weak.

How Users Drive Value in Two-sided Markets: Platform Designs that Matter

Extant research has popularized the perspective that strong network effects produce “winner-take-all” outcomes. This leads platforms to invest in user growth and encourages investors to subsidize them. However, without user stickiness, strong network effects in the current period may fade in future periods, so a user growth strategy is not effective. By adding a time dimension to the network effects, we develop a model of cross-period and within-period network effects to explain how different types of network effects drive value. We emphasize that the cross-period same-side network effect contributes to user stickiness while the within-period cross-side network effect persists to be conditional on user stickiness. We theorize that one reason platforms have heterogeneous cross-period same-side network effects is because of the product learning mechanism: products with higher uncertainty can expect a higher cross-period same-side network effect. Using Groupon data, we verify our theoretical insight and discuss platform design choices that enhance user stickiness when the cross-period same-side network effect is weak.

Analysis of Network Effect Enhancement Method of Short-video Platform under the Background of MCN Mechanism Popularization - Taking Douyin as an Example

Recently, with the rapid development of Internet entertainment, Chinas Multi-Channel Network markets (MCN) has been expanding, and MCN have significant advantages in many aspects. Meanwhile, Douyin has also maintained a forward development trend since its launch. Therefore, the research topic is to explore how to enhance Network Effects of short video platform on Douyin in the context of the extensive MCN. During the research process, the necessity and scientific nature of Douyin + MCN is analyzed by SWOT analysis, and it is concluded that Douyin + MCN is a Positive Trans-Boundary Network Effect Model, which accords with the law of Winner-takes-all. Finally, countermeasures are proposed for the future challenges of Douyin + MCN. Therefore, in order to further expand Network Effect, Douyin should make use of MCN to achieve high-quality video content, seize the dominant position, control the content direction, appropriately subsidize MCN, balance the market development, encourage the development of small and medium-sized MCN, and avoid the situation of one or several dominant MCN. Besides, it is suggested to create a better creation environment, attract more users, and ultimately achieve the purpose of maximizing Network Effects.

General Additive Network Effect Models

In the interest of business innovation, social network companies often carry out experiments to test product changes and new ideas. In such experiments, users are typically assigned to one of two experimental conditions with some outcome of interest observed and compared. In this setting, the outcome of one user may be influenced by not only the condition to which they are assigned but also the conditions of other users via their network connections. This challenges classical experimental design and analysis methodologies and requires specialized methods. We introduce the general additive network effect (GANE) model, which encompasses many existing outcome models in the literature under a unified model-based framework. The model is both interpretable and flexible in modeling the treatment effect as well as the network influence. We show that (quasi) maximum likelihood estimators are consistent and asymptotically normal for a family of model specifications. Quantities of interest such as the global treatment effect are defined and expressed as functions of the GANE model parameters, and hence inference can be carried out using likelihood theory. We further propose the “power-degree” (POW-DEG) specification of the GANE model. The performance of POW-DEG and other specifications of the GANE model are investigated via simulations. Under model misspecification, the POW-DEG specification appears to work well. Finally, we study the characteristics of good experimental designs for the POW-DEG specification. We find that graph-cluster randomization and balanced designs are not necessarily optimal for precise estimation of the global treatment effect, indicating the need for alternative design strategies.

Adoption of additive manufacturing: A survey of the role of knowledge networks and maturity in small and medium-sized Danish production firms

Additive manufacturing (AM) has had a significant impact on manufacturing processes in many industries, which is expected to increase much further in the coming years. AM is, however, not easy to implement, as this involves several knowledge-related challenges, particularly for small and medium-sized enterprises (SMEs). To add to the understanding of this phenomenon, we develop a theoretical model that describes the use of external networks for overcoming such challenges in part by acquiring AM knowledge to increase internal AM maturity. We test our model of direct and mediated network effects on AM-related competitive advantages through a survey of Danish manufacturing SMEs. The results show that increased AM maturity explains around 40 percent of the relationship between knowledge acquisition from networks and AM-related competitive advantages. Results also show that building internal AM maturity was especially driven by networking with AM developers, AM suppliers, and AM organizations, while networking with AM consultants and other AM users promoted competitive advantages through other mechanisms. This suggests two overall strategies for achieving competitive advantages from AM: (1) using networks to build AM maturity, and (2) using external AM resources to compensate for low internal AM maturity.

Maujan.pk: Nothing Ventured; Nothing Gained

The case helps the students explore a capacity utilization problem faced by the CEO of Maujan.pk. The financial statements from the fiscal year 2013–2014 put the CEO, Tahir, in a difficult spot. According to these reports, his main business, which was online discount coupons, has been outpaced in growth by his side business, Facebook management. Tahir had launched the Facebook management business as an afterthought; his strategic business decisions never factored it as a major consideration. Now, Tahir is forced to reflect on this matter, especially because of resource constraints. Being a small company with six employees, he has had to share resources between the two businesses. Now that his side business has grown faster than the main business, Tahir finds himself understaffed. Meanwhile, the original business is slowing down, and few avenues of growth are forthcoming due to network congestion (network effects model). Keeping this in mind, a decision must be made to choose one of the two businesses or expand and continue them both. The expansion would mean significant logistic overhead and initial resource starvation. This is perplexing for someone like Tahir, who is a real hands-on manager. Eventually, inaction would lead him to these problems, as the business would continue to grow to further compete for the same resources. Each decision has its merits and demerits in a fast-moving Pakistani digital market where copycat businesses can spring up within days, and customer loyalty is mercurial at best.

Estimating the correlation in network disturbance models

AbstractThe network disturbance model of P. Doreian (1989), expresses the dependency between observations taken at the vertices of a network by modelling the correlation between neighbouring vertices, using a single correlation parameter $\rho$. It has been observed that estimation of $\rho$ in dense graphs, using the method of maximum likelihood, leads to results that can be both biased and very unstable. In this article, we sketch why this is the case, showing that the variability cannot be avoided, no matter how large the network. We also propose a more intuitive estimator of $\rho$, which shows little bias. The related network effects model is briefly discussed.

Optimal Block Designs for Experiments on Networks

Abstract We propose a method for constructing optimal block designs for experiments on networks. The response model for a given network interference structure extends the linear network effects model to incorporate blocks. The optimality criteria are chosen to reflect the experimental objectives and an exchange algorithm is used to search across the design space for obtaining an efficient design when an exhaustive search is not possible. Our interest lies in estimating the direct comparisons among treatments, in the presence of nuisance network effects that stem from the underlying network interference structure governing the experimental units, or in the network effects themselves. Comparisons of optimal designs under different models, including the standard treatment models, are examined by comparing the variance and bias of treatment effect estimators. We also suggest a way of defining blocks, while taking into account the interrelations of groups of experimental units within a network, using spectral clustering techniques to achieve optimal modularity. We expect connected units within closed-form communities to behave similarly to an external stimulus. We provide evidence that our approach can lead to efficiency gains over conventional designs such as randomised designs that ignore the network structure and we illustrate its usefulness for experiments on networks.

A Model for Joint Engineering-Economic Analysis of Cellular Systems

Economic models of network effects, which describe how the market for a good or service depends on the number of units sold, have been used for decades in the economics literature. However, it is not clear how these models may be applied to the cellular network service. The relationship between the “value” of cellular service and the “size” of the cellular network (including the base station density, user density, and bandwidth), is not straightforward. Furthermore, this relationship is dependent on physical and technical characteristics of the network. This letter develops a joint engineering-economic model for network effects in cellular networks. The model captures the relationship between the network size and its technical characteristics, and the behavior of the market for that service. It is shown how the developed model can be used to find equilibrium size of the market for different networks. For example, this model identifies and formalizes fundamental differences between the markets for millimeter wave and microwave networks, that are not revealed by standard engineering or economic analyses.

Study of bilateral trading platform competition in consideration of suppliers’ spatial network

The bilateral trading platform is an intermediary system which serves both consumers and suppliers. To make this happen, platforms must attract users on board from both sides and make them trade with each other. Spatial distribution of suppliers has the significant impact on consumers platform adoption decision, but how does the supplier’s network affect the platform competition has rarely been studied. In this paper, the supplier’s spatial network is added to bilateral trading platform network effect model, and influences of the supplier’s network structure on the platform competition are discussed. Simulation results show that, compared with the central network, the decentralized network is more conducive to the platform competition. When suppliers on two competitive platforms follow the same network structure, the reinforcement effect of the first-mover advantage on platforms will vanish. This result is robust in different spatial network patterns. Conversely, when one of the platforms has a superior network structure of suppliers, its first-mover advantage will get strengthened, and it will achieve a better competition result; otherwise, its first-mover advantage will get weakened. These results provide an insight into the platform competition mechanism and will help platforms to make corresponding competitive strategies.

Cortical functional hyperconnectivity in a mouse model of depression and selective network effects of ketamine.

See Huang and Liston (doi:10.1093/awx166) for a scientific commentary on this article.Human depression is associated with glutamatergic dysfunction and alterations in resting state network activity. However, the indirect nature of human in vivo glutamate and activity assessments obscures mechanistic details. Using the chronic social defeat mouse model of depression, we determine how mesoscale glutamatergic networks are altered after chronic stress, and in response to the rapid acting antidepressant, ketamine. Transgenic mice (Ai85) expressing iGluSnFR (a recombinant protein sensor) permitted real-time in vivo selective characterization of extracellular glutamate and longitudinal imaging of mesoscale cortical glutamatergic functional circuits. Mice underwent chronic social defeat or a control condition, while spontaneous cortical activity was longitudinally sampled. After chronic social defeat, we observed network-wide glutamate functional hyperconnectivity in defeated animals, which was confirmed with voltage sensitive dye imaging in an independent cohort. Subanaesthetic ketamine has unique effects in defeated animals. Acutely, subanaesthetic ketamine induces large global cortical glutamate transients in defeated animals, and an elevated subanaesthetic dose resulted in sustained global increase in cortical glutamate. Local cortical inhibition of glutamate transporters in naïve mice given ketamine produced a similar extracellular glutamate phenotype, with both glutamate transients and a dose-dependent accumulation of glutamate. Twenty-four hours after ketamine, normalization of depressive-like behaviour in defeated animals was accompanied by reduced glutamate functional connectivity strength. Altered glutamate functional connectivity in this animal model confirms the central role of glutamate dynamics as well as network-wide changes after chronic stress and in response to ketamine.

A discrete choice framework for modeling and forecasting the adoption and diffusion of new transportation services

Major technological and infrastructural changes over the next decades, such as the introduction of autonomous vehicles, implementation of mileage-based fees, carsharing and ridesharing are expected to have a profound impact on lifestyles and travel behavior. Current travel demand models are unable to predict long-range trends in travel behavior as they do not entail a mechanism that projects membership and market share of new modes of transport (Uber, Lyft, etc.). We propose integrating discrete choice and technology adoption models to address the aforementioned issue. In order to do so, we build on the formulation of discrete mixture models and specifically Latent Class Choice Models (LCCMs), which were integrated with a network effect model. The network effect model quantifies the impact of the spatial/network effect of the new technology on the utility of adoption. We adopted a confirmatory approach to estimating our dynamic LCCM based on findings from the technology diffusion literature that focus on defining two distinct types of adopters: innovator/early adopters and imitators. LCCMs allow for heterogeneity in the utility of adoption for the various market segments i.e. innovators/early adopters, imitators and non-adopters. We make use of revealed preference (RP) time series data from a one-way carsharing system in a major city in the United States to estimate model parameters. The data entails a complete set of member enrollment for the carsharing service for a time period of 2.5years after being launched. Consistent with the technology diffusion literature, our model identifies three latent classes whose utility of adoption have a well-defined set of preferences that are significant and behaviorally consistent. The technology adoption model predicts the probability that a certain individual will adopt the service at a certain time period, and is explained by social influences, network effect, socio-demographics and level-of-service attributes. Finally, the model was calibrated and then used to forecast adoption of the carsharing system for potential investment strategy scenarios. A couple of takeaways from the adoption forecasts were: (1) placing a new station/pod for the carsharing system outside a major technology firm induces the highest expected increase in the monthly number of adopters; and (2) no significant difference in the expected number of monthly adopters for the downtown region will exist between having a station or on-street parking.

New market development of platform ecosystems: A case study of the Nintendo Wii

The platform ecosystem could generate unlimited innovation with outside complementors and various consumers. Some researchers focused on the competition between two platform ecosystems in the same market and constructed theoretical models of indirect network effects between complementors and consumers. However, the implications from these researches might not adapt the platform ecosystem, which aims to explore new markets. This study focuses on a platform ecosystem—the Nintendo Wii—that explores a new market in the Japanese video game industry. We analyze why Wii rapidly declined in spite of its significant initial installed base. Our results show that the strength of indirect network effects for complementors' software provision became relatively small in the case of competition between new and existing markets. Software provision of complementors and software selection of consumers led to the concentration of consumer segments in the new market. Software purchases of consumers in the new markets were dependent on the provider's name and on seasonality and increased the risk for complementors. Consequently, complementors exited from the platform ecosystem of the Wii new market; as a result, more consumers could not participate in the platform ecosystem. Based on the findings, this study makes suggestions for platform ecosystem success.

Optimal Design of Experiments on Connected Units with Application to Social Networks

SummaryWhen experiments are performed on social networks, it is difficult to justify the usual assumption of treatment–unit additivity, owing to the connections between actors in the network. We investigate how connections between experimental units affect the design of experiments on those experimental units. Specifically, where we have unstructured treatments, whose effects propagate according to a linear network effects model which we introduce, we show that optimal designs are no longer necessarily balanced; we further demonstrate how experiments which do not take a network effect into account can lead to much higher variance than necessary and/or a large bias. We show the use of this methodology in a very wide range of experiments in agricultural trials, and crossover trials, as well as experiments on connected individuals in a social network.

The ‘ripple effect’ of schedule control: a social network approach

This study extends research on workplace flexibility by integrating insights from literature on social networks. This approach adds to our understanding of the meso-level social context in which workplace flexibility is enacted and highlights the importance of coworkers. Drawing on social information processing theory and literature on flexibility inequity and inducement effects, I investigate whether an individual's schedule control is associated not only with his or her own job attitudes, as has been established in past literature, but also with the job attitudes of his or her coworkers. Using a network data set comprised of 337 individuals from one firm and employing a network effects model, I examine four types of possible coworker effects and ultimately find evidence of an effect among job network peers, such that individuals whose job network peers have greater schedule control report lower job satisfaction and organizational commitment. I conclude with implications for research and practice that deepen our understanding of the social context of flexibility.

Network structure effects on incumbency advantage

Research summary: The literature on network effects has implicitly assumed that an increase in the size of the installed base magnifies network effects, which is a source of incumbency advantage. We argue that the overemphasis on this relationship has resulted in controversy and confusion in the literature, where the role of social networks remains largely unaddressed. By developing computational models of network effects with various network structures, we show that social distance in a customer network plays a moderating role that strengthens or weakens the relationship between the installed base and network effects, which in turn, affects the durability of incumbency advantage. When the average social distance between members in a customer network is large, the incumbency advantage will not be amplified, and an entrant with an incompatible product or service may find ways into the market. On the other hand, when the average social distance is small, early entry with a growing installed base will magnify incumbency advantage.Managerial summary: In evaluating the strength of incumbency advantage or determining the price of an early mover, the size of the installed base has been widely used. We find that it is not a sufficient statistic, and confusion and error appear to result from assuming that it is. Our study suggests that degrees of separation, a measure of social distance in a network, can provide managers with an additional yardstick to sharpen their evaluation. When customer networks are characterized by fewer degrees of separation, the conventional use of the installed base as a metric may be reasonable. On the other hand, when customer networks are characterized by larger degrees of separation, the conventional use may potentially mislead managers in their decision‐making. Thinking about the roots of user benefits (e.g., access to a few significant others vs. hubs) may be a reasonable starting point for assessing degrees of separation in a customer network. Copyright © 2015 John Wiley & Sons, Ltd.

Examining the effect of social influence on student performance through network autocorrelation models

The paper investigates the link between student relations and their performances at university. A social influence mechanism is hypothesized as individuals adjusting their own behaviors to those of others with whom they are connected. This contribution explores the effect of peers on a real network formed by a cohort of students enrolled at a graduate level in an Italian University. Specifically, by adopting a network effects model, the relation between interpersonal networks and university performance is evaluated assuming that student performance is related to the performance of the other students belonging to the same group. By controlling for individual covariates, the network results show informal contacts, based on mutual interests and goals, are related to performance, while formal groups formed temporarily by the instructor have no such effect.

Substitutes, Complements and Network Effects in Instant Messaging Services

Discerning the behavior and the mechanisms underlying network effects remains an elusive but economically important research objective. This research provides insight into these mechanisms through an investigation of the growth of network effects, and their modulation through substitutes and complementary services offered by China Mobile. Our research shows that network effects can be perceived as a manifestation of the consumer adoption life cycle behaviors, and that contagion models provide an accurate description of the mechanism underlying network effects. In the period under study, China Mobile’s customers adopted personal computer base instant messaging at a declining rate, abandoning it due to cannibalization by competitor services. In contrast short message service instant messaging on low cost platforms demonstrated the strongest network effects and traditional early adoption ‘innovation’ consumer behavior was unimportant in defining China Mobile’s instant messaging markets. In those markets, network effects arose from growth of the own service network, but were strongly influenced by cannibalization from network effects from close substitutes or complements. The research develops an accurate causal model of the impact of network size on network adoption rate. The model of network effects, complements and substitutes presented in this paper can be used to predict specific consequences of managerial decisions to increase network size through promotions, merging and acquisition of networks.

Application of Network Autocorrelation Models in Psychological Studies: Taking the Impact of Peer Effect, Learning Motivation on Adolescents' Academic Performance as Examples

用常规的线性模型来处理彼此之间并不独立的关系数据，会违背线性模型关于独立性的假设。引入网络自相关模型，在介绍其发展及原理的基础上，通过模拟研究比较网络效应模型和常规线性模型在处理关系数据时的差异。结果表明，在关系数据的参数估计和拟合程度上，网络效应模型都显著优于常规线性模型，宜采用网络效应模型来处理关系数据。最后以同群效应、学习动机对青少年学业表现的影响为例，探讨该模型在心理学研究中应用的可能性和必要性。