The Development of IT Service Quality Standards in China

Abstract

The process of developing IT-based standards has recently being labeled “a critical research frontier for IS research”. Within the last half-dozen years, IS scholars have begun to examine a variety of domains in which standards-setting occurred. Many of these studies have focused on international telecommunications standards in Asia. Despite the number and diversity of the many empirical studies, most have focused on technical standards for ICT interoperability. We consider the creation and dissemination of a distinct type of standard, known as “IT service management standards”, which are a type of process standards specific to IT vendors. To our knowledge, the subject of establishing service measurement standards has not been considered in the IS literature. We review the growing literature on standards-setting and standards-adoption, and we inquire whether the issues that consistently accompany the development process for technical standards for IT and telecommunications interoperability will generalize to the process of developing IT service measurement standards. We describe a case study of the development of standards for measuring IT services in China. Rather than analyzing our data according to a specific theory, we compare the results to date of China’s creation of IT service standards to the phenomena frequently observed in prior IS studies of standards-setting. We argue that several recurring phenomena that accompany prior standards-setting initiatives in the IS literature (i.e., conflict among parties, warring factions, etc.) are lacking in our study. We speculate about the reasons why conflict is absent and develop proposition that can be evaluated with future case data.

Appendix: Company Feedback from IT Vendor Firms in Shanghai and Hubei (Translated from Mandarin to English by the Authors)

Companies	Comments and suggestions
	Comments from Shanghai vendors
Yitong international	In the dimension of reliability, there is overlap and correlation among several indicators: the rate of serious incidents, the rate of continuous service operation, and the rate of major security incidents. Maybe it can be measured by the client’s degree of satisfaction with the SLA. However, according to our situation, many clients do not define the SLA in the contract clearly, which will make it difficult for the overall assessment
	The ratio of services delivered upon the first requests may not be an appropriate indicator. Although most service requests could just be solved in the telephone, as the user access the hotline, but it is not always the case. Some problems that are not emergent and cannot be solved immediately can be recorded and forwarded to experts. It should be a more appropriate way to notify the user when the problem is solved
	The rate of service report submitted also depends on the agreement with the user and the customer’s type of business. Telecommunications companies, for example, never submitted a user service report, which is not directly related to its quality of service
	Several indicators are not cleared defined (e.g., corporate brand, qualification in industry, market share). There are no clear criteria that can be referred
	There are some indicators associated with whether the service norms are established, but the applicability of the norms is not taken in account
Yatai	In the security dimension, the ratio of unauthorized data changes to the total number of data changes is calculated as an indicator. The problem is how to record the change events and ensure the authenticity of the records
Hanwei consulting	It is suggested to include the rate of repeated failures solved as an indicator of reliability. It is defined as the ratio of the number of the same (repeat) requests to the total service requests
Puhua	It will be good to develop some industry-specific standards, so that the measurement of the service quality will be more applicable
Tianji	Requires quantitative evaluation of the terms, such as marked quantitative content, conducive to enterprise with reference to specific data validation and assessment
Wanda	The service quality model should be adaptable to be fit for situations of different companies
Nantian	There are several indicators on compliance, which means whether the services are achieved based on corresponding norms and standards. In fact, these indicators are of little significance, but also difficult to operationalize, because in many cases the appropriate norms or standards are not available
	The indicator of the ratio of requests processed is actually difficult to operate, because usually we can only record the number of service requests that we received—not the number of requests that we do not complete. It is suggested to delete this indicator
Shanghai 30wish	It is too idealistic to try to build general quality standards for various types of IT services. In fact, it is also not conducive to wide adoption of IT services standards in the future. It is recommended that the working group should develop different service quality measurement for various types of IT services
	Comments from Hubei vendors (south-central China)
Xingde	Those indicators on compliance are very difficult to understand and operationalize. It’s not clear which norms and standards should be referred
	For the dimension of security, it is not clear whether all the data owned by a client is considered, or if it only refers to the data provided by the vendor
	There are too many indicators in the service quality evaluation model. It needs to be customized in order to fit the different types of IT services
	There is paradox between the qualitative nature of the index and the quantitative metrics. For example, the “completeness of function” is very difficult to quantify according to a service catalog. In practical operation, we can replace completeness with reliability, which can be measured by the ratio of timely responses or effective solutions to the number of such requests
Wuhan 30wish	It’s very difficult to develop a general IT service quality measurement system, because different types of IT services will vary greatly in their content and delivery mode. Even for the same general type of IT service (IT operations and maintenance services), they can be divided into different service subclass with different quality requirements
	For some indicators related to our IT operation and maintenance practices, we may have different requirement, sometimes better than the standards. For example, as to the dimension of tangibles, we require not only deliverable documentation stage by stage, but also visible evidence of service process on ITSM platform
	Some indicators are not clearly defined and even unreasonable, and there are some overlaps between several items
	Some metrics are too idealistic or complex to be used in actual practice. For example, the rate of confidentiality accidents does not make sense because these kinds of accidents are usually never allowed in service agreements
Navimentum	The service quality model is too versatile. In the service delivery process, even the services of operation and maintenance for software applications are different from services of operation and maintenance for hardware system. Although the model can be modified according to different service categories. It doesn’t give appropriate guidance on how to adapt the model. If it can be modified arbitrarily, then what’s the significance and value of the standard?
	Some of the indicators are very difficult to use for operations. Many items should be defined in the service agreement, but in the actual situation, many contracts are very simple and do not include these data. It’s therefore difficult to evaluate the service quality
Lilosoft	In the actual situation, the different IT services have different characteristics. It’s necessarily to modify the measurement system to evaluate different services. There are no specific guidelines on how to modify and choose the dimension and metrics
Sunflower	It is not reasonable to simply assume that there is a generic IT service quality measurement model. Some relevant standards such as ITIL, ISO20000, etc. are all process based to define the quality of services. There is no general method to quantify the service quality or evaluate the service level
Dongrun	The overall service quality model is relatively reasonable, but the metrics should be more specific and detailed. It should be adaptable to be fit for the features of different types of services, otherwise it would be difficult to use in the actual operation