Application Portfolio Management: APM in a multi-actor context

Application Portfolio Management: APM in a multi-actor context

Van Luipen, B.

Wijers, G.M. (mentor)
Bouwman, W.A.G.A. (mentor)

Technology, Policy and Management

Sectie Informatie en Communicatie Technologie

2009-12-07

An application is a specific class of information systems that supports the business processes of an organisation directly (Riemp et al., 2007). Many companies and organisations rely on applications to operate and support their daily operations. The complete set of applications of an organisation is called an application portfolio. As time goes by the application portfolio of an organisation grows and becomes more complex in order to support the changing business needs. As a result maintenance and operational costs grow and the adaptation of the applications to support the business becomes more difficult due to the complexity (Fabriek et al., 2007). Application Portfolio Management (APM) is a framework to structure the decision making for the investments in the application portfolio and should result in a reduction of maintenance cost and complexity, but also in an optimal support of the business processes. The APM framework defines the processes for the decision making and couples the business strategy to the application architecture in order to ensure the support of the business processes (Eck et al., 2004). Rijkswaterstaat (RWS) has centralised all IT facilities including the applications to the DICT, the IT department at RWS. The DICT encounters major problems to manage their application portfolios. The DICT is overwhelmed with hundreds of existing applications with the corresponding requests and wishes of the business, but the DICT simply does not have the capacity to fulfil all the requests of the business. In addition many actors from different departments with different goals and ideas are involved in the current APM decision making processes. All these actors have also problems with the current APM processes, but perceive also the outcomes of the APM processes different. The situation at RWS is recognised as a multi actor context, because the actors have different opinions and goals, but are dependent on each other to accomplish their goals. The APM theory only describes the theoretical processes for optimal decision making about the application portfolio, but does not take a multi actor-context into account. In the research project, process management is selected as additional theory and combined with the APM theory to solve the problems at RWS. Process management is a decision making process that considers the organization as a network of actors that have to solve a common problem. Process management typically applies in a network of actors that is not hierarchically organized, but exists of more or less autonomous actors that have different goals, but are dependent on each other to reach those goals (de Bruijn et al., 2007). The strategy to solve the APM problems is that the involved actors start a process in order to solve the APM problems together. In the research project an actor analysis is applied to establish a provisory agenda of APM issues and the actor configurations necessary to solve these issues. An inventory of possible conflicts among the actors, recognised as dilemmas, that might block the process is also provided. The main issues on the agenda are the absence of a business architecture and the absence of the selection process. The business architecture is required to derive the desired functionality and application architecture from the business strategy. The selection process is required to select the most valuable applications and most important functionality within boundaries of budget and capacity. Many of the problems the DICT originate from the absence of the selection process and business architecture. The actual design of the process, the negotiation of the final agenda and process rules among the involved actors and the actual process execution are not part of the research project. RWS can decide to start the actual process to improve the APM processes based on the proposed agenda, actor configuration and the dilemmas. This is also a major limitation of the research project. Only the outcomes of the actor analysis are validated by the client of the research, but the proposed strategy to start a process is not validated at all. Additional research is needed to validate the process management approach to resolve the APM related problems. Another weakness is the chosen research approach. The case study at RWS only provides results for the specific situation at RWS. The agenda, the actor configurations and the dilemmas all apply to the situation at RWS and do not apply to other organisations. However the link between the missing APM processes and some issues on the agenda, some actor configurations and some dilemmas are potentially more general than the situation at RWS only. Besides the limitations of the research project the process management approach appears to be useful to analyse the multi-actor context of the APM processes. Process management relates the problems the regarding to APM to actors configuration and potential conflicts among actors and provides a course of action for an acceptable solution for the involved actors.

Application Portfolio Management
IT Management
Process Management

http://resolver.tudelft.nl/uuid:9387cc4a-cea3-4d77-860e-6218fa3c7cae

2009-12-07

Campus only

Student theses

master thesis

(c) 2009 Van Luipen, B.