Can enterprise architectures reduce failure in development projects?

Abstract

Purpose Scant attention has been given to the role of enterprise architecture (EA) in relationship to risk management in information system development projects. Even less attention has been given to the inter‐organizational setting. The aim of this paper is to better understand this relationship. Design/methodology/approach The relationship between EA and project failure/success is investigated by – through a workshop – creating a retrospective view on the use of architectures in large and complex ICT‐projects. Findings Failure factors can be grouped in organization network, people, process, product and technology categories. The findings show that a disappointingly limited number of public sector development projects make sufficient use of architecture as a risk management instrument. Architectures should be considered both as a risk‐mitigating instrument and as an organizational shaping routine to reduce project failure and manage risk in organization networks. Research limitations/implications A single workshop with a limited number of participants was conducted. The findings need further refinement and generalization based on more empirical research investigating the relationship between architecture and project failure. Practical implications Architecture should give explicit consideration to risk management and help to draw attention to this. Governance mechanisms need be defined to ensure that the organizations' members become aware of both architecture and risk management. Risk management and EA have similarities, as they are both an instrument and an organizational shaping routine. Originality/value Governments collaborate more and more in organizational networks and for that reason often multiple organizations are involved in information system project developments. Enterprise architecture as a risk mitigation instrument has not, to date, been given attention.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 32 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.