EVALUATING AND REFINING THE ‘ ENTERPRISE ARCHITECTURE AS STRATEGY ’ APPROACH AND ARTEFACTS

Enterprise Architecture (EA) is a new discipline that has emerged from the need to create a holistic view of an enterprise, and thereby to discover business/IT integration and alignment opportunities across enterprise structures. Previous EA value propositions that merely focus on IT cost reductions will no longer convince management to invest in EA. Today, EA should enable business strategy in the organisation to create value. This resides in the ability to do enterprise optimisation through process standardisation and integration. In order to do this, a new approach is required to integrate EA into the strategy planning process of the organisation. This article explores the use of three key artefacts – operating models, core diagrams, and an operating maturity assessment as defined by Ross, Weill & Robertson [1] – as the basis of this new approach. Action research is applied to a research group to obtain qualitative feedback on the practicality of the artefacts.


INTRODUCTION
Contrary to the information technology and cost reduction foci of previous EA endeavours, this research is used to emphasise a new value-creation focus that includes business architecture and enables business strategy.In support of this new focus, Ross et al. [1] defined a new EA approach that incorporates EA decision-making as part of the strategic decision-making processes of an organisation.Action research is used to gain qualitative feedback on the perceived practicality of two key artefacts that are used to underpin this new approach.

ENTERPRISE ARCHITECTURE DEFINED
The first traces of EA were found in the publication of Zachman [2].Zachman [3] defined EA as follows: "Descriptive representations (i.e.models) that are relevant for describing an enterprise such that it can be produced to management's requirements (quality) and maintained over the period of its useful life (change)."Zachman introduced the Zachman framework, which consists of various models that are used to define and communicate six characteristics/abstractions (What, How, Where, Who, When, and Why) for five different viewpoints/perspectives (Planner, Owner, Designer, Builder, and Sub-contractor) (Zachman [3]).The Zachman framework "is a tool for managing and communicating the vast amount of information needed to make broad decisions, those that enable the organisation to be competitive" (O'Rourke, Fishman & Selkow [4]).
Numerous EA definitions were formulated following the inception of the Zachman framework.These definitions addressed the following elements with different emphases:


Providing a systems view -i.e.describing systems, their components (e.g.people, processes, information, and technology), their interaction, and interrelationships.This includes the use of decomposition strategies to ensure holistic solutions in terms of solution components (TOGAF [5]; Theuerkorn [6]; Gartner in Lapkin [7]; Handler [8]).


Providing a blueprint for directing the company in terms of required high-level processes and IT capabilities (Ross et al. [1]; Gartner in Lapkin [7]; Boar [9]). Defining a process / master plan to explore and model the current realities and the envisioned future state, and enable its evolution (Gartner in Lapkin [7]; Bernard [10]; Schekkerman [11]).

ENTERPRISE ARCHITECTURE -CREATING GOVERNANCE ON A STRATEGIC LEVEL
EA initially aimed at modelling / describing the architecture components associated with information technology.EA value was limited to direct improvements in the performance of IT itself (lowering overall costs from IT).This approach demonstrated some form of return on investment (ROI) -i.e.accelerating project start-up and decreasing investment in staff, consulting, training, and tools.
Today, EA has broadened from enterprise-wide IT architecture (EWITA) to include business architecture (BA); that is, EA = BA + EWITA (Malan & Bredemeyer [14]; Bernard [10]; Ross et al. [1]).The focus is on optimisation "across boundaries to achieve system goals" and the "translation of strategy into implementation" (Malan & Bredemeyer [14]).The change in focus is closely related to the restricted contribution of previous EA value propositions.EA practitioners realised that EA could show more significant value when used to improve business performance, and with IT used to support the execution of strategy (Rosser [15]; Lapkin [16]).

Research design
Action research was selected for qualitative research for the following reasons:


The 'EA as strategy' approach of Ross et al. is still new (published in 2006).Research respondents needed to have a good understanding of EA in general and of the new 'EA as strategy' approach.The 'Business Architecture' post-graduate course was used as a vehicle to convey knowledge about EA and the 'EA as strategy' approach, techniques, and artefacts to students, who were then used as respondents.


The action research process provided the opportunity to assess the students' understanding of the course content, and guide them towards the correct use of the 'EA as strategy' approach, techniques, and artefacts.
The action research process that was followed is based on the work of specialists (referred to by Hodgkinson and Maree [22]):  Planning -A literature study was conducted in the field of EA to design the course content and assessment mechanisms.Special emphasis was placed on strategic management, the 'EA as strategy' approach, techniques and artefacts, the business architecture domain, and the development of an EA plan. Implementation -Live presentations from the course presenter and industry speakers, course notes, and literature references were used to convey the course content to students.Students then had the opportunity to work individually or in pairs and to select an organisation in which to implement some of the techniques presented in the course.An interim project report was submitted for assessment.Students also wrote a semester test to assess their understanding of EA principles and of the 'EA as strategy' approach, techniques, and artefacts defined by Ross et al. [1].


Observation -The course presenter observed/assessed the students' understanding of the course content.Feedback was given to the students in the light of their semester test and interim project report.Students now had the opportunity to improve/update their project reports and submit a final project report.Based on the final report, they had to submit a completed survey. Evaluation -The final reports were assessed and surveys were analysed.Analysis of qualitative survey feedback gave new insight into the practicality of two key artefacts (operating models and core diagrams).New insights were used to define suggested improvements, recommendations, and an agenda for further research.
The survey consisted of twenty-eight questions.Some of the questions were taken from the on-line survey used by the Institute for Enterprise Architecture Developments (IFEAD) (Schekkerman [23]).Categorisation of business activities was taken from the Oracle Magazine subscription form (Oracle Magazine [24]).Questions were categorised according to parameters that could have a significant influence on the perceived practicality of defining the two key artefacts -the operating model and the core diagram (see Figure 6).

SUMMARY AND INTERPRETATION OF RESULTS
It was found that most of the students had an engineering background, held positions related to business process planning and improvement, and showed significant interest in the management / improvement of business processes and organisational management / governance.Students also had sufficient knowledge of information systems.
Concerning the organisation profile, most of the companies that were used for analysis purposes employed fewer than 10 000 employees, and were involved in a large number of business activities excluding the financial sector.Results further indicated a relatively low level of operating maturity -most of the analysed companies displayed business silo behaviour, while none of the companies operated according to a modular business design.
The study indicated that business architecture was well established at the analysed companies.Use of architecture modelling technology was limited.
The perceived practicality of the operating model and core diagram artefacts could not be evaluated on a corporate level, as most of the students defined operating models at a business unit level.According to Ross et al. [1], this should not be a hurdle in validating the artefacts per se, as operating models and core diagrams may be defined at various levels of the organisation.The interpretation of the various difficulties experienced follows:  Difficulty in selecting a single operating model is linked to the identification of the degree of process standardisation / integration for the analysed organisation / business unit.Extensive implicit/explicit knowledge is implied during the evaluation of the operating model characteristics that define the degree of process standardisation / integration. Students had difficulty in finding the correct information to perform an operating model classification or select core diagram components.Identification of operating model characteristics and core diagram components requires knowledge about the strategic choices (markets, products/services), operating/organising logic, business processes, and main databases and technologies of the organisation.Some baseline architectures are thus required, and this knowledge is not necessarily available or in an explicit format.


Students experienced difficulty in selecting the main components of the core diagram and understanding the core diagram templates.This may be related to the limited set of examples provided in the textbook.Case studies would be required to demonstrate inputs that would be required (e.g.baseline architectures) to define the core diagram components.

CRITICAL EVALUATIONS AND INFERENCES
Based on the qualitative feedback received from the action research effort, the researcher revisited the main objectives of the operating model and core diagram: To aid the main stakeholders / users of these artefacts (business and IT managers) in guiding them during their strategic decision-making processes.


To communicate architecture vision to other stakeholders (in terms of process standardisation / integration requirements).
If the main stakeholders are to use these artefacts to guide them during the strategic decision-making processes, the artefacts should be based on a more rigorous approach to attaining the artefact outputs.This will increase their validity and reliability.The researcher also believes that process standardisation / integration requirements should be based on a more scientific approach to define optimal standardisation / integration requirements for an organisation.Porter [26], for instance, believes that decisions regarding process standardisation / integration are complex and require detailed analysis based on the strategic intent of the organisation (e.g.cost leadership / differentiation / focus-driven for target segments).Cost leadership companies, for instance, would have to assess the impact that process standardisation / integration could have on overall cost, while differentiation-focused companies need to assess if process standardisation / integration could increase the uniqueness of an activity or lower its cost of differentiation.

CONCLUSIONS
This study emphasised the limited value gained from EA when measured in terms of ROI due to cost reductions alone.Today EA practitioners realise that new value propositions emerge when EA is used to support the strategic direction of the organisation.This new focus was used to introduce a new approach towards EA value creation, called 'EA as strategy'.The approach incorporates EA planning as part of the strategic decision-making process using three key artefacts: operating models, core diagrams, and an operating maturity assessment.
Action research was used to assess the practicality of two key artefacts (operating model and core diagram), which highlighted some difficulties that were experienced and led to some critical evaluations and recommendations regarding the artefacts.It is believed that the operating model and core diagram could be useful in visualising the process standardisation / integration requirements of an organisation / sub-division.The artefacts should, however, be supported by a more scientific approach to their derivation, to increase their validity/reliability.
Further research has been initiated to perform a case study at an organisation.The case study incorporates processes to model baseline architectures, current strategic choices (markets, products/services), operating/organising logic, business processes, main databases, and technologies of the organisation.This will be followed by various analyses (e.g.value chain analyses) to identify process standardisation / integration opportunities.Current artefact designs (e.g.operating model and core diagram) may need to be adapted to convey the process standardisation / integration requirements to strategic decisionmakers.The new artefact designs will be distributed to different strategic decision-makers to gain feedback about their usefulness during strategic decision-making.

Figure 6 :
Figure 6: Parameters that influence the practicality of defining two key artefacts