Chapter 9
The Science of Service Systems and Networks
Manufacturing dominated the global economy during the last couple of centuries. Both academics and practitioners thus paid significant attention to the design, development, production, and innovation of physical products. With their contributions to the development of manufacturing science and technology, the manufacturing industry has considerably improved its production productivity and the quality of made products. In the second half of the twentieth century, in particular, the world witnessed a long period of prosperity in all aspects of well-being that were mainly driven by the spread of industrialization and substantially increased manufacturing productivities around the world.
Today, the quality of life has taken into account not only the material standard of living but also other intangible values of living that are recognized to be mainly service-oriented. As discussed in Chapter 2, the global economy has shifted its focus from manufacturing to services to meet the changing needs of human beings. Indeed, entering the information era has accelerated the shift, which created unfilled gaps in the service science and technology. Indeed, service organizations have been on the hunt for appropriate methodologies and tools that can help them engineer and manage their service offering and delivery throughout the service lifecycle at the scale they would like to reach, efficiently, cost-effectively, and globally (Spohrer and Riechen, (2006); Qiu, (2012).
As discussed earlier, the effectiveness (
) of a service as a solution to meet the changing needs of customers is equal to the product of the quality (
) of the technical attributes of the solution and the acceptance (
) of that solution by the customers, that is, 
. However, the acceptance of customers changes rapidly, varying with time, places, cultures, and service contexts. Because people's acceptance is largely subjective, manufacturing mindsets with a focus on physical attributes indeed become ineffective when applied in the field of service engineering and management. Hence, to address the discussed change acceleration phenomena with scientific rigor, we must develop service science based on people-centric and service mindsets.
Promisingly, the introduction of putting employee and customers first in 1990s made the first breakthrough in developing people-centric and service mindsets. Since then, service organizations have begun to develop, operate, and manage businesses and measure their successes by focusing on both customers' satisfaction and employees' job satisfaction, resulting in an operational philosophy shift in business operations and management. This book essentially presented such a new perspective of service study. We took a holistic view of the service lifecycle to explore the dynamics of service systems and the structure and behavior of people-centered service networks.
By defining service as a cocreation transformation process enabled and executed by a service system, we discussed how the performance of the service system could be quantitatively analyzed using a holistic approach. By leveraging the advances in computing and network technologies, social science, management science, and other relevant fields, we demonstrated that service networks in light of service encounters could be comprehensively explored in a closed-loop and real-time manner. The presented science of service should help service organizations understand and capture market trends, design and engineer service products and delivery networks, operate service operations, and control and manage the service lifecycles for competitive advantage.
In this final chapter, we first summarize this book by providing some final thoughts on the development of service science in a comprehensive manner. We strongly advocate that the service research and practice community must appreciate and continue to develop a variety of methodologies and tools that can be well derived and evolved from the well-known theories and principles in systems theory, operations research, marketing science, organizational behavior and theory, network theory, social computing, and analytics. In Section 9.2, we then conclude this chapter by articulating that innovative approaches to the development of service science are truly on demand. The science of service will be well developed by the scholars and practitioners worldwide in an evolutionary and collective manner.
9.1 The Science of Service Systems and Networks
Holistically, a service organization is a service system, essentially consisting of service providers, customers, products, and processes. As compared to a producing-goods system, a service system must be people-centered. Therefore, a service system surely is sociotechnical. On the basis of the earlier discussion, we understand that it is the transformation process that ties all other system constituents together and cocreates the respective values for both service providers and customers. Whether the values can be fully met relies on the efficient, effective, and smart business operations, which must be engineered, executed, and managed with scientific rigor across the service system.
Now it is crystal clear that service is people-centric, truly cultural and bilateral. The type and nature of a service dictates how a service is performed, which accordingly determine how a series of service encounters could occur throughout its service lifecycle. The type, order, frequency, timing, time, efficiency, and effectiveness of the series of service encounters throughout the service lifecycles determine the quality of services perceived by customers who purchase and consume the services. On the basis of the discussions in the preceding chapters, we understand that people-centered, interactive, and behavioral activities in a service system essentially engender a service interaction cocreation network or simply service network. Indeed, as the velocity of globalization accelerates, the changes and influences are more ambient, quick, and substantial, impacting us as providers or customers in dynamic and complex ways that have not seen before. The understanding of service networks becomes essential for service providers to be able to design, offer, and manage services for competitive advantage.
Because of the sociotechnical nature of a service system, we use a systems approach to evaluate the performance of the service system, aimed at capturing both utilitarian functions and sociopsychological needs that characterize service systems. However, the true people's behavioral and attitudinal dynamics of a sociotechnical system requires performing real-time social network analytics. As a result, the insights of service interactions in the formed service networks can be truly explored and understood, which assist stakeholders to make respective while cooperative informed decisions at the point of need to improve their service cocreation processes across the service lifecycles in an optimal manner.
Bearing the earlier discussion in mind, we consider a service as a transformation process rather than simply an offered service product. Truly, both provider-side and customer-side people are always involved in an interactive manner, directly or indirectly and physically or virtually, throughout the transformation process. Hence, we view a service as a value cocreation process. For a service, goods are frequently the conduits of service provision; the physical attributes and technical characteristics that specify the goods are indispensable to the service. The quality (
) of the technical attributes in the service, indeed, mainly defines the quality of the goods. To a service customer, 
is frequently perceived in service provision as the quality of designated service functionalities that are defined in a service specification. As described in the equation of 
, the value of 
also directly depends on the value of 
, which is largely related to sociopsychological perceptions of the customer throughout the service lifecycle. 
is subjective in nature, varying with people, time, places, cultures, and service contexts (Figure 9.1).
Figure 9.1 Engineering and managing competitive services: holistic and lifecycle perspective.
Service is highly heterogeneous. Each service is unique as a unique customer and a service provider agent essentially cocreate the service values that meet the respective needs of the customer and the service provider. The variability of service and the need for measuring sociopsychological perceptions had made extremely challenging the exploration of the service lifecycle, which spans market discovery, engineering, delivery, and sustaining, in an integrated and holistic manner. Figure 9.1 highlights a holistic and lifecycle viewpoint of how we should engineer and manage competitive services in the twenty-first century.
It is well understood that the science of service is essential for a service organization to achieve the ultimate goal of engineering and managing competitive services in its service marketplace. As discussed earlier in this book, the prior lack of means to monitor and capture people's dynamics throughout the service lifecycle has prohibited us from gaining insights into the service engineering and management in a service organization for years. However, we believe that the convergence of the following advances in science and technology has made possible the design and development of the needed methods and tools that can facilitate service organization to monitor and capture people's dynamics throughout the service lifecycle:
- Digitalization
- Networks and telecommunications
- Collaborative methods and tools
- The fast advances in social network media
- Big data technologies and analytics methods and tools
Figure 9.2 shows how in a systems and operations perspective a service organization can be successively and real-time transformed for competitive advantage by fully leveraging the convergence of the above-mentioned advances in science and technology.
Figure 9.2 Engineering and managing competitive services: systems and operations perspective.
“People-centric sensing will help drive this trend by enabling a different way to sense, learn, visualize, and share information about ourselves, friends, communities, the way we live, and the world we live in” (Campbell et al., (2008). From the discussions in the preceding chapters, we understand that voluminous, real-time, and heterogeneous data on the service cocreation dynamics of both service providers and customers can be comprehensively captured and analyzed if service systems are well planned, designed, and operated as illustrated in Figure 9.2. In other words, when the enabling technologies are appropriately implemented, we can surely create and execute smarter working and consuming practices so that we can make service cocreation processes not only beneficial but also enjoyable. As a result, services are competitive and satisfactory.
Because of the enablement of people sensing and computational thinking with the support of the above-mentioned advances in science and technology, enormous opportunities truly lie ahead of us. However, if the science of service is not well developed, we cannot ensure that service systems will perform in such a way that the respective values for both service providers and customers can be optimally met, at present as well as in the long run. By leveraging both systems methods and networks analytics, in this book we essentially present one promising approach to develop the needed methods and tools, making a contribution to the body of knowledge in service science (Figure 9.3).
Figure 9.3 Engineering and managing competitive services: scientific perspective.
As shown in Figure 9.3, we advocate that a combined systems and network approach can help service organizations engineer and manage their competitive services. The presented approach in this book fundamentally focuses on identifying actionable areas for service improvements across all service system constituents in a holistic, comprehensive while cost-effective and efficient manner. The presented framework is integrative, quantitative, and closed-loop in nature. As a result, a service system with cocreation processes can be modeled, explored, monitored, and controlled with scientific rigor.
Innovatively, systems and network approaches are integrated in this book. When combined and applied to the field of service engineering and management, they are complementary. A systems approach to gain the fundamental understanding of how a service system as a whole behaves must be first investigated. Specifically, we apply structural equation models (SEM) to describe the systems' performance and/or conduct necessary hypothesis testing and/or confirmatory factor analyses. Secondly, social and collaborative network approaches to explore the interactions and insights of people-centered service networks can be employed. For instance, we apply social network analysis (SNA) models to explore how service networks across the service lifecycle are formed and behaved and understand how the service networks might evolve over time. Consequently, we as service providers can always make optimal decisions at the point of need, strategically and tactically, so that we can innovate, market, engineer, and execute services in a competitive and satisfactory manner.
9.1.1 Enhancing the Approaches to Explore Service Systems and Networks
As explained earlier, we present one approach to develop the needed methods and tools through leveraging the strengths of both system methods and network analytics. Indeed, this could be a perfect starting point for us to analyze the systems' behavior, the network structures, and dynamics of a sociotechnical service system. In the preface, we articulate that this book focuses on the development of a real-time and closed-loop framework to help service organizations engineer and manage their service systems. That is to say, developing an approach to model service systems while allowing performing continual improvements is surely unique, differentiating this book from others. However, we truly understand that the presented approach can be further enhanced and developed. More importantly, we are sure that there must be many other approaches to develop the science of service.
Regardless of the variability of services and the complexity and heterogeneity of service systems, the discovery, design, engineering, and delivery of services must be fully supported by the science of service if service organizations wish to stay competitive from time to time. In general, the foci of decision-making change with the mix of 8 Ps that substantially varies with the progression of service offering and delivery. Therefore, depending on circumstances, we have to customize and further enhance known approaches to explore service systems and networks. As a matter of fact, we must frequently develop new approaches to engineer and manage services in order to meet the needs of service providers and customers over time.
We can take the simplified model illustrated in Figure 8.6 as an example. When significant variations of online classes exist, an SEM based on the prior knowledge might be substantially deviated from the reality. To ensure that we can validate the SEM, we have to find an appropriate way to enhance the modeling. For instance, we could apply the probabilistic-based analysis methodology such as SEM-based and semisupervised Bayesian networks to the exploration of class-dependent collaborations, which might help to improve the accuracy of analyses if significant variations of online classes do exist.
Generally speaking, an identified best practice can be effectively adopted as a general guideline by a service organization in its daily service engineering and managerial operations. However, certain ongoing changes must be applied in the process of service offering and execution for optimal outcomes as each service is unique. Hence, the service industry is looking for practical and scientific service engineering and management approaches that can be applied in a gradual and evolutionary manner. Ultimately, the framework proposed in Figure 9.3 should be fully implemented in a real-time and closed-loop manner, which is graphically illustrated in Figure 9.2. Indeed, there is a long way to go in the service academia and industry before the science of service gets well developed. A full exploration of the science of service is surely necessary in both the academia and industry. A brief discussion in this regard is provided in Section 9.2.
9.1.2 A Pragmatic Approach to Explore Service Systems
On second thought, if people are not the focus of a study in a service system, an alternative approach to explore service systems might be more appropriate than one illustrated in Figure 9.3. This is particularly true when a practically applicable transition in service operations and management is crucial for the time being for a service organization to survive in a fiercely challenging and competitive marketplace. In other words, by applying well-known methods and tools to explore and address the ongoing changes in the marketplaces, service organizations can make swift and appropriate changes and actions to transform operations and management in an evolutionary manner so that they can continue to engineer and execute quality and satisfactory services to meet the needs of their customers.
For example, the performance of a service system is frequently related to business units' operational efficiencies from a managerial perspective. If a study of business units' operational efficiencies is indeed critical for a service organization at a given business period, well-known methods and tools can be practically adopted. For instance, we can take advantage of the following modeling technologies, analytical hierarchy process (AHP), data envelopment analysis (DEA), principal component analysis (PCA), and partial least squares (PLS), to collectively study the dynamics of service systems with a focus on exploring service operations and management on the service provider's behalf. Figure 9.4 illustrates how AHP, DEA, PCA, and PLS can be seamlessly and integratively applied in support of this alternative investigation. Note that this alternative approach highlights a viewpoint of pragmatism as rich data on operational functions and decisions in a service organization are most likely available at present.
Figure 9.4 An integrated approach for improving daily service business operations.
As shown in Figure 9.4, AHP is typically used for comparing a list of objectives or alternatives where the problem elements are structured in an organizational hierarchy. AHP depends on experts' knowledge to provide rankings that eventually lead to weights being assigned for the relative importance of different categories defined based on the problem elements and targeted organizational hierarchy. These weights are then input to a DEA procedure to determine corporate and business units' efficiencies. PCA and PLS are strictly data-driven modeling technologies. By leveraging these previously mentioned data-driven technologies and expert-based operations exploratory models, decision makers can gain insights into service systems and hence operate them in a competitive manner.
In practice, PLS methods can facilitate the identification of operational weakness, which considerably relies on the output of the DEA analysis as well as other available systems dynamics data (Figure 9.4). However, these data could conceivably be exceedingly large, and much of it is probably of little value in generating an analytical model. It is crucial to use variables that truly affect the analytical output of the model. Data that does not influence the exploration only serves to degrade the performance of the model and needs to be eliminated.
In the integrated model shown in Figure 9.4, genetic algorithms can be used to select meaningful variables. From the potential candidate variables, a process of random selection of variables can be used to generate a set of models forming an initial population. The choice of variables is defined by a binary word with a one in the bit corresponding to a variable used in the model, and a zero in the position of the variable not used. Next, this population is evaluated to obtain an estimate of the standard error of prediction for each member. Those models with low values of the standard error are better than those with higher ones. The models are rank-ordered from lowest to highest standard error of prediction. A probability can be then assigned to each model that is inversely proportional to the standard error of prediction. Two models are chosen at random with the probability of selection equal to the assigned probability of the model to be used in a breeding process to produce the next generation. One or two random integers are chosen from one to the number of variables. These integers define the crossover points that are used with the binary words to define the next generation of models. The crossover points define where the binary words defining the two models chosen are broken. The broken pieces are then rejoined to define the set of variables to be used in the next generation of models. In addition, some of the bits in the binary are randomly flipped, representing a mutation. In this manner, a new generation is produced. Sometimes, a small fraction of the best models from the previous generation is carried forward so that if an exceptional model is generated, that model is not lost by the breeding process. This procedure is repeated until a stopping criterion, such as a certain standard error of prediction or the number of generation, is reached.
As a result, these selected variables are used to generate a PLS model. The model can be further used in a manner described earlier to help service systems improve the performance. However, this approach does not take into account that there is a resource availability issue (e.g., cost) involved in any improvement scheme. Eventually, an approach that seeks to optimize this approach such as determination of the lowest cost approach to improve profits by a fixed percentage. The chosen variables may not directly provide the lowest cost to make improvements. Using PCA, we can learn the structure of the variables and with this information, remove certain variables and replace them with other variables that might provide a solution with a lower cost. Alternatively, the genetic algorithm might be used to find replacement variables that are less costly. From a set of candidate variables, the genetic algorithm would search to find those variables that are the best predictors. This search can be guided by PCA by choosing those variables related to more expensive ones.
In summary, this alternative approach could help service organizations evaluate, compare, and optimize service business operations when they are facing severe competition in the presence of massive uncertainty and risk in their operating environments. The ultimate goal of this approach is to help service organizations transform their practices for competitive advantage with the support of the following well-developed analytical scenarios in sequence:
- AHP depends on existing algorithms and/or new inputs from experts to provide the knowledge for weights assignment for the relative importance of different input/output variables in the organizational and operational hierarchy.
- The AHP output provides weights used by DEA to generate the organization's operational efficiency. Apparently, the quantified outputs combined with those identified weak areas in service business operations better help the service organization understand where they stand in competition and what they could address in improving their performance in terms of operational efficiency.
- Genetic algorithms can be employed in preprocessing the inputs. Through PCA, the structure of the variables is learned. With a better understanding of the circumstances, on one hand, certain variables can be removed; on the other hand, the identified variable correlations can be utilized in facilitating the prediction generations of quantities of the primary interest in the next step.
- PLS can then be used to generate predictions of quantities of primary interest under the circumstances. The primary interest, for example, can be profit, throughput, or more sophisticated definable systems outcomes.
- Comparisons of generated predictions can be conducted through sensitivity analysis by selecting highly influential input variables. When facing massive uncertainty, this integrated model can be utilized in quantifying the consequences when different transformations in operational practices could occur under different circumstances, assisting management in making informed decisions (e.g., a series of optimal changes or transformation actions) to improve systems performance while minimizing potential risks.
9.2 The Science of Service in the Twenty-First Century
Generally speaking, best practices in service engineering and management in the service industry can be effectively adopted as operational and managerial guidelines by service organizations to support and manage their daily operations and business activities. Because each service is unique, it is necessary for both the service providing-side people and consuming-side customers to cocreate the respective values of services in a practical, viable, and competitive manner. To make this happen in a satisfactory manner in both the short term and the long run, the framework illustrated in Figure 9.3 must be well incorporated into the service lifecycle shown in Figure 9.1. As a result, a service organization, with the support of effective service engineering and management that is enabled in a real-time and closed-loop manner shown in Figure 9.2, can offer and deliver competitive services throughout the service lifecycle.
The science of service is still in its early infancy stage although it emerged in the early 2000s (Qiu, (2012). Without question, a well-defined and more developed service science would better facilitate service organizations in conducting service engineering and management across service value-added networks. In reality, capable and competitive service systems must be highly adaptable and sustainable to their service environment (when, where and who to deliver, and whom to be served). Therefore, the developed science of service must span all service offering and delivery areas from engineering and/or managing service marketing, conceiving, design, quality assurance, regulatory compliance, operations, to innovation throughout the lifecycle of service.
Regardless of methods and tools that can be utilized at each stage of the service lifecycle, meeting the needs of people at the point of need is what actually matters in operating competitive service systems. People involved in service are unique and truly different from each other, for example, individuals as customers who have different needs, individuals on the service provider side who are assigned with certain roles and responsibilities, managers who are in charge of designated business domains, executives who are overseeing service organizations, and collaborators who are contributing to the service offering and delivery networks. To ensure that each service of a service system can be well executed, people involved in service must collaborate with each other well throughout the service lifecycle, which can be accomplished only if four interdependent and essential flows (Qiu, (2013) in support of the service system and formed service networks are engineered and managed with scientific rigor (Figure 9.5).
Figure 9.5 Four interdependent and essential flows in support of service systems and networks.
Let us start with the customer dynamics flow. We know that meeting both the utilitarian and psychological needs of customers by focusing on a chain of interactive service encounters is the key to explore the customer dynamics flow. Hence, the customer dynamics flow must be explored with the support of behavioral science, consumer behavior and dynamics, and cognitive science. Understanding the customer dynamics flow becomes essential for service organizations to capture market trends and get ready for and capable of offering and delivering excellent customer experience.
The organizational behavior flow plays a key role in forming functional service networks. The organizational behavior flow focuses on organizational capability development and competence alignment in support of meeting the customers' utilitarian and psychological needs. Organizational behavior flows must ensure that service organizations can offer and deliver a chain of interactive and positive service encounters while realizing a competitive level of employees' job satisfaction. Therefore, the organizational behavior flow must be explored with the support of behavioral science, cognitive science, individual and group dynamics, organizational dynamics, operations management, and workforce optimization, making sure that service organizations can continuously improve their job satisfaction and organizational behavior.
The physical flow focuses on the conduits of service provision. An efficient and effective physical flow can provide employees and customers with the right tools, servicescape, and other necessary resource supports to facilitate service encounters in meeting both utilitarian and psychological needs of customers while improving job satisfaction. In today's information era, the effectiveness and efficiency of a physical flow considerably rely on the effectiveness and efficiency of a corresponding information flow. An information flow must capture right data/information in a timely manner and then support the operational and managerial needs of employees and customers in an intelligent way across the service lifecycle. An optimal information flow shall promptly enable the right data, information, and knowledge service for employees and customers at the point of need.
This book takes an innovative and unique approach to contributing to the development of service science. Under a given circumstance, one step at a time, we explore a unique research area in a given service context. Collectively, the service research community must explore the defined four flows across service systems in a comprehensive and holistic way. Theoretically, the dynamics of service systems in terms of both systems performance and service networks behavior must be fully explored, understood, and controlled so that the respective values for service providers and customers can be optimally cocreated. In practice, it simply becomes how different methods and tools can be made available at the point of need in real time so that individuals, managers, executives, and collaborators can interactively, effectively, and collectively perform their responsibilities and duties in the processes of transformation in meeting their respective needs.
Once again, we now fully recognize people as the focus during the service production and consumption process in service provision. We learn that different people have their personal traits in the physiological and psychological perspectives, different cognitive abilities, and unique sociological constraints. It has been exceedingly challenging for the service research and practice community to investigate methods and tools that can be well applied for modeling and exploring people's behaviors in service because people-sensing mechanisms in service were hardly enabled not long ago. The recent and fast advances in sensor-based networks, pervasive and mobile computing, online social media, and big data methodologies and tools indeed have changed this (Figure 9.6). Therefore, we are sure that it is time for the service research and practice community to develop service theories and principles that can be applied in effectively managing and controlling systemic behavior, leveraging sociotechnical effects, and stimulating innovations throughout the service lifecycle (marketing, design and engineering, operations, delivery, benchmarking, and optimization for improvement).
Figure 9.6 A holistic and integrated approach to contributing to the development of service science.
Without question, advanced descriptive, predictive, and prescriptive service science studies surely rely on the continual development of systems theory, operations research, management science, marketing science, advanced computing and communication technology, network theory, social computing, and analytics. As a matter of fact, the science of service as a metascience of service must build on predecessors' excellent work from many of the above-mentioned disciplines (Larson, (2011); Qiu, (2012). However, a variety of innovative approaches for the development of the science of service are truly on demand in today's global service-led economy. This book takes an innovative and unique approach to contributing to the development of service science (Figure 9.6). Specifically, we take a holistic view of the service lifecycle and explore the real-time dynamics of service systems and networks.
In conclusion, the service industry is in need of descriptive, predictive, and prescriptive research of service in a holistic, integral, and quantitative manner. There is a marvelous Chinese saying, “cast away a brick and attract a jade stone.” Hopefully, this book serves such a purpose. We are confident that the science of service will be well developed by the scholars and practitioners worldwide in an evolutionary and collective manner. Ultimately, the developed body of knowledge and tools in this emerging interdisciplinary field can be effectively applied by service organization to address their service challenges in the twenty-first century's service-led economy.
References
- Campbell, A., Lane, N, Miluzzo, E., Peterson, R., Lu, H., Zheng, X., Musolesi, M., Fodor, K., Eisenman, S., & Ahn, G. (2008). The rise of people-centric sensing. IEEE Internet Computing, July–August, 12–21.
- Larson, R. (2011) Foreword in Service Systems Implementation, ed. by H. Demirkan, J. Spohrer, and V. Krishna. Springer.
- Qiu, R. G. (2012). Editorial column—launching service science. Service Science, 4(1), 1–3.
- Qiu, R. G. (2013). We must rethink service encounters. Service Science, 5(1), 1–3.
- Spohrer, J., & Riechen, D. (2006). Services science. Communications of the ACM, 49(7), 30–34.