Project Compliance with Enterprise Architecture

Abstract

This research project set out to identify effective practices and models for working with projects that are required to comply with Enterprise Architecture (EA), and investigate the benefits and drawbacks brought about by compliance. Research methods used are canonical action research, a statistical survey study (n=293), focus groups and a structured literature review. Concrete contributions include a descriptive framework of different types of architecture (such as enterprise, domain and project architecture), best practices and an artifact and process model for projects conforming to EA (targeted at practitioners), a conceptualization of compliance (including four dimensions and compliance checks), a statistical explanatory model for EA conformance and benefits, and a typology of compliance tactics. In addition, the research has contributed to our knowledge on Enterprise Architecture and compliance in general. We will shortly elaborate on this below. The statistical study showed that most of the identified techniques are used in practice. Important compliance tactics in this respect are: knowledge exchanges, providing advice, working with a PSA, conducting compliance assessments, and having management propagate EA. Financial sanctions and disincentives proved to be barely used in practice. The results also showed that projects benefit in several ways from working with EA, namely delivering the desired quality more often, being better equipped to deal with risk and being able to manage complexity more effectively. Several of the hypothesized project benefits could not be confirmed, however, such as exceeding deadlines and budgets less often, and delivering the required functionality more frequently. EA simply did not have a significant effect on these aspects. The results also demonstrated that EA has a downright negative impact on the speed of initializing projects, probably due to EA introducing additional project complexity (e.g. getting acquainted with abstract EA prescriptions, dealing with additional stakeholders and balancing possible conflicts of interests). The results furthermore showed that several organization-wide benefits are achieved due to EA, namely accomplishing enterprise goals, gaining insight into organizational complexity, achieving integration, standardization and deduplication of related processes and systems, depicting a clear image of the future situation and providing a communicational frame of reference. Hypothesized organizational benefits not confirmed are achieving business/IT alignment and agility, controlling costs and complexity, and co-operating with other organizations effectively and efficiently. The statistical results furthermore showed that, although most techniques identified are used in practice, only three have a significant impact on achieving compliance with EA: compliance assessments, management propagation and assisting projects. Compliance of projects subsequently results in an increased ability to achieve EA-related benefits. At the organizational level, conformance has significant effects on achieving business/IT alignment, accomplishing enterprise-wide goals and integrating, standardizing and deduplicating processes and systems. At the project level, conformance is shown to have significant effects on the ability to manage project complexity, and delivering the desired quality and functionality. Interestingly, project compliance with EA has the strongest effects on organization-wide benefits, whereas projects themselves benefit to a lesser extent and in more subtle ways. In any case, the project level is shown to play an important part in achieving organization-wide goals, which is one of the key goals of EA.