Model of design principles for the definition of data for the consistency of evaluation in Enterprise Governance of IT

Various current developments in society and the technical possibilities in the economy have created a new era. The age of digitalization. It is not only about the change from analogue to digital, that would be digitisation. The term digitalization refers to the complete conversion of countless business models. The term is to be settled thereby both within the range of the information technology in addition, in the business management. The new digital business models show completely new aspects of organizational theory and information technology. Technically, the economy has never been so networked as it is today. This offers the possibility of new ways of providing services. The next stage of evolution in industrialization has thus been reached, the socalled Industry 4.0 which permits new value chains. The framework conditions for an enterprise are difficult to define and this applies also to the border of the own enterprise. Governance structures play an important role here. In order to ensure that these structures are presented transparently, companies also rely on data. Data that is no longer collected internally, but originates from big data structures. One no longer speaks of a few sources but of a socalled data lakes. In the last years, many researches have dedicated their work in IT Governance and numerous methods, techniques and tools were proposed. In the meantime new digital business models developed rapidly and aligning business with IT have become more difficult ever. Usually organisations manage themselves with several governance frameworks next to IT Governance frameworks. Different governance frameworks are related, dependent and interact under each other and since they have all their own method of measurement and evaluation, the interaction suffers to a lack of data consistency or they do not take all dimensions of decision making into account. Furthermore the lack of consistency is caused by definition of governance frameworks itself and lead to concurring goals among them. This dissertationthesis has the main aim to search for a model of design principles of data definition for consistency of evaluation in Enterprise Governance of IT (EGIT) by applying a multi method approach which has been undertaken in a design science research (DSR) project. The dissertation project follows various goals such as (1) determine the state of the art dealing with governance in the digital age, (2) provide a set of data principles based on rigor science proof and quantitative evidence, (3) design the model of principles for the definition of data,(4)contribute to science with an artefact of a set of principles definition of data, and finally (5) contribute the results to broader audience. The relevance of this consistency problem has been proven by undertaking several design research actions in real circumstances such as in (1) a project in the banking sector (major Swiss bank) focussing on IT Governance in 2016, (2) a study in the hospital sector (Swiss hospital) focussing on IT Governance in the context of an Bring Your Own Device (BYOD) approach in 2017, (3) a digital readiness survey with approximately 260 distributed questionnaires among Swiss companies with a result of a dataset (n=67) in spring/summer 2017 and with (4) a study in the banking sector (Swiss and international banks of different size) focussing on Budgeting methods for IT project in fast fast changing market conditions in 2018 and finally, (5) a study for the designing of principles of data definition for the use of measuring governance have been undertaken to build the artefact of this dissertation project. In addition, the rigorous analysis of existing theory have been proven by (1) critically review the existing state of art of measurement and evaluation concepts of governance frameworks by following a comprehensive review of the normative literature dealing with the measurement aspect of IT Governance and (2) taking part in studies with the topic of digital business models or IT and DataGovernance respectively. With the exception of the first project with the major Swiss Bank, these studies have been presented on various conferences. Large organizations suffer from an organizational history that is incompatible with rapidly changing governance structures. Reorganizations lead to different levels of current perception of an organization. The result is an inconsistent flood of information for decision making, which, however, is not accepted by decisionmakers as a targetoriented data basis. The BYOD study showed that it is very difficult to maintain IT and data governance in a dual business model (simultaneous B2B and B2C). Furthermore, the evaluation of reliable system usage data is very difficult in a datacritical environment of patient data in a hospital. But BYOD is indispensable. The study on the digital maturity of Swiss companies showed that the management of IT digitization is much more critical than the general management. The perception of digitization in IT departments differs greatly from that of business departments. The overwhelming majority of business representatives believe they have a digital strategy where exponents of the IT departments express themselves far away from it. The study also showed that Swiss companies are strongly processoriented and believe in their processes. A finding that was examined in more detail in the last study using the example of a budget process for IT projects. The latest study showed that generally speaking, most interviewees are satisfied with their current, mostly traditionally kind of IT project budgeting process. But the processes faces some challenges such as coordination and structural challenges but also poor business IT alignment and a lack of flexibility. This lack is seen in other factors than processes such as culture and the structure of the organization that seem to be much more important when it comes to the ability to react to changing markets. The literature review showed that models in this area of research strongly focus on IT Governance and DataGovernance as such. Studies are more strongly focussed on revealing business effects of implemented IT Governances on performance. Syntactic, semantic, and temporal circumstances can lead to inconsistencies in data that users perceive as poor data quality. It is in the nature of data that the content itself has no choice but to become inconsistent sooner or later. Literature tries to overcome this circumstance by defining a socalled data governance. In the context of data warehousing, ETL processes are defined which carry out a socalled datacleansing as a step of data refinement. In most cases, however, data inconsistency can only be partially eliminated and the basis for decisionmaking remains unsatisfactory. Finally, the study Principles of Data Definition for the Use of Measuring Governance showed a consistent picture of the principles. Experts agreed that the elements for the metrics of data quality can be categorized into the variables content, function and coherence. The field study among the companies surveyed confirmed this finding. The hypothesis could not be rejected, however, only to variable content which leads to the conclusions. Data that is needed for a longer period of time must be equipped with a kind of levers, which helps to process the contents syntactically, semantically and temporally correctly at any time. At the same time, it must always be possible to determine whether the claimed data (i.e. the content) is the last content claimed to be correct. There are countless possibilities to define a data content and therefore it is extremely difficult to define a single standard. This dissertation thesis therefore aims at proposing a model for the principles of data definition to be followed to ensure a consistent evaluation of a governance framework. The results of the survey on the model of principles showed that there is agreement among experts on the existence of these principles and confirmed the model. However, this could not be confirmed in the business context. The model is an abstract solution. Inpractice, only the verification of the correctness of the content of data seems to be important. Neither a semantic quality check nor syntactical correctness of the processing were considered important by the respondents. This can be interpreted as a misunderstanding. It is to be expected that tha model can be used as a starting point for any data refinement process. However, it is developed within the framework of EGIT and limitations cannot be excluded from this circumstance. It should also benoted that the study was mainly carried out in Europe. The expert interviews were conducted exclusively in Switzerland. Switzerland is regarded as a country with a high level of data processing maturity. Thanks to Switzerland’s large, rigorously regulated financial sector, the criticality of data quality is well developed. Both the survey of experts and the field study should therefore not be regarded as generally valid. This is because the field study could only be carried out with participants who are already intensively involved in the subject of dataquality in a professional manner. However, due to the results and the awareness of the framework conditions for good dataquality, it must be assumed that the actual deficits with regard to data quality in data processing that is not checked for quality are considerably greater than among the respondents to the field study. When asked about syntax and semantics, respondents confirmed a high correlation to the corresponding items. Only the relationship to the perceived data quality is not recognized. This obvious misunderstanding can certainly be seen in companies as a reason for a lack of data quality and could be investigated in further studies.