Atos Blog

Enterprise architecture – Is it worth investment?

July 23rd, 2009 Atos Blog Tags: enterprise architecture, Facebook, PRINCE2, TOGAF, Twitter
Posted in CIO Agenda, Strategy & Innovation, Technology, Trends |

The current recession is causing most organisations to rethink IT expenditure and significantly cut back on costs; for example, the UK government’s Operational Efficiency Programme requires Departments to save £3.2bn per year in total IT costs in four years time compared to current expenditure, a 20% reduction.  Yet 2009 also marks the release of the TOGAF 9 framework, a significant milestone in the maturity of enterprise architecture.  It is therefore pertinent for CIOs to ask whether enterprise architecture is worth investment at present.

But first, let’s define what we mean by enterprise architecture.  An enterprise is a group of one or more related organisations, departments or units, and could also include the ‘extended enterprise’ of suppliers, partners and customers.  An enterprise architecture is a formal description of the structure and inter-relationship of components making up the enterprise.  The enterprise architecture activity is “a means of creating a coherent way of modeling an enterprise to enable the efficient and effective deployment of IT”, according to John Zachman, who originated the concept in the 1980s.

An enterprise architecture includes a set of architectures typically covering business, data, applications and technology, where each is described as a set of component lists, matrices, diagrams and related standards.  These architectures might have several versions: a current version, a future target version and one or more incremental transitional versions – the scope and level of detail of each will normally vary depending on their relevance to business issues.

So why do enterprise architecture? Organisations primarily seek improved IT efficiency, which comes from lower IT procurement, development and maintenance costs through the sharing and re-use of business analysis, data, software and technology products.

Service levels should be more closely matched to business needs through traceable links from business to technology components.  Open infrastructure service and communication standards should facilitate more portable and interoperable applications that are easier to upgrade and so more responsive to changing business needs.  IT governance decisions should be more objective if they are based on an explicit architecture.

Moreover, enterprise architecture should also reduce the risk of major change programmes through more coherent and consistent IT deliverables coming from conformance to common standards.  Better analysis of strategic change options – such as shared services, mergers and major cost reduction – using linked, traceable architectural components will also reduce risk and accelerate results.

The 1996 Klinger-Cohen Act mandated that US Federal IT programmes must utilise an enterprise architecture framework as part of their IT capital planning and investment control process, so it is now widely used in the US public sector.  Enterprise architecture has been adopted widely in Europe: for example, by the Department for Work and Pensions and the Ministry of Defence in UK central government; and by Shell, BNP Paribas, Novartis, KLM, Visa, Axa, Aviva, and Yorkshire Water in the private sector.

But why then isn’t enterprise architecture even more widespread? Well firstly, the benefits can be hard to understand, harder to cost-justify and even harder to track since they are often only realised in long-term improvements in IT efficiency and in the avoidance of problems in change programmes.  Enterprise architecture is not normally a ‘quick win’ so it can lack management attention and be vulnerable to cuts in downturns.  CIOs therefore need to clearly sell the benefits of enterprise architecture, and ensure that they are tracked and regularly realised.

Some enterprise architecture initiatives founder because they become too introverted; they get lost in content, they neglect project and management users, and they fail to express results in language understandable to users.   As a result, they acquire an academic, ivory tower image and became irrelevant to business issues, making them vulnerable to the next round of cost-cutting.  Architects need to engage with, and stay relevant to the projects that use their architectures and the senior management that make strategic change decisions. They should do this by prioritising their work around the most serious business and IT issues, and planning their deliverables to regularly deliver useful results.

An enterprise architecture is worthless if it is ignored by projects going their own technical way.  Architects can, in isolation, struggle to assure and enforce architectural compliance when they lack influence and project deadlines are threatened.  CIOs should therefore establish and integrate architectural governance into their overall IT governance processes so that enforceable compliance decisions can be made that take into account all relevant factors.

Managing the complexity of enterprise architecture whilst still making progress will be a challenge for most organisations.  Architects will need to select an appropriate architectural framework and customise it rather than inventing their own method.  They will need to invest in tool support and they should explicitly scope their deliverables to only go to the level of detail needed to resolve business and IT issues.  They may also decide to partition complex architectures into two or more simpler ones with limited linkages.

What is the future for enterprise architecture? Well, the recession will slow the rate of adoption in the short-term as organisations cut discretionary costs.  However, in the longer term, enterprise architecture will continue to be adopted in large, complex organisations undergoing rapid rate of change, particularly where the following are important:

• the need to shared data across organisational boundaries;
• service oriented architecture implementation; and
• the increasing competitive importance of business and IT alignment.

TOGAF has acquired so much industry-wide commitment and content in its latest version 9 that it will become the de facto architectural framework standard in the same way as PRINCE2 has in project management, probably leading to the eventual absorption of expensive-to-maintain proprietary architectural approaches like Capgemini’s Integrated Architecture Framework and IBM’s SOMA.

The increasing adoption of Cloud computing business services will continue to move enterprise architecture away from a technical infrastructure architecture focus as infrastructure services become relatively less important.  Instead enterprise architecture will move towards information systems, and particularly application services, with the steady uptake of service oriented architectures.

The business adoption of popular web 2.0 online collaboration applications like Facebook and Twitter will require information architectures for handling unstructured information held as text, image and audio as well as structured data.  However enterprise architecture should be able to incorporate these and other likely future challenges to consolidate its position as an important part of corporate IT planning and management.