Enterprise architecture management_s_impact_on_information_technology


Published on

Published in: Technology, Business, Education
1 Like
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Enterprise architecture management_s_impact_on_information_technology

  1. 1. Proceedings of the 44th Hawaii International Conference on System Sciences - 2011 Enterprise Architecture Management’s Impact on Information Technology Success Robert Lagerström, Teodor Sommestad, Markus Buschle, Mathias Ekstedt Industrial Information and Control Systems The Royal Institute of Technology Osquldas väg 12, 10044, Stockholm Sweden {robertl,teodors,markusb,mek101}@ics.kth.se Abstract encompassing models. Some use enterpriseBoth practitioners and researchers put forward architecture primarily for documenting the currententerprise architecture management as a mean for state of affairs, whereas others see it as planning andachieving success with information technology. Many decision making support. Yet others see enterprise architecture mainly as a means for communication.arguments have been put forward to support the Many, of course, also use EA for more than one ofbenefits claimed to arise from mature enterprise these different purposes. The general belief amongarchitecture management and a considerable amount both practitioners and researchers is that EA surelyof literature describes the components of mature can enhance the present and future performance of(successful) enterprise architecture management. the business, by satisfying several needs for severalHowever, few studies have empirically tested whether stakeholders in the organization.the enterprise architecture management activities A line of criticism that has been directed mainly from practitioners towards EA during the last coupleimpact organizations’ success with information of years is that it is a very heavy approach. There aretechnology. This paper tests the relationship between lots of steps that need to be taken, if the state of theorganizations’ success with information technology art methods are followed and there is an extensiveand enterprise architecture management activities. amount of models to be developed according to manySignificant correlations are found between these frameworks. In many cases there seem to be anvariables. impression that enterprise architecture is failing to deliver the promised benefits since they cannot be delivered quickly and cheaply enough to the organization. This criticism could for instance be1. Introduction traced in the EA communities trying to become more lightweight [4][5]. Even larger more all Enterprise Architecture (EA) has during the last encompassing frameworks such as the TOGAF [6]two decades been put forward as an effective are also making very clear explicit statements that theapproach for business and information systems framework needs to be tailored before it is applied.management [1]. Support for tasks such as product Lately, a web 2.0 inspired approach is starting toplanning, business development, and business grow [7]. Gartner on their side predicts that EA hasprocess- and information systems consolidation are just past the bottom of the “Through offor instance suggested to be offered by enterprise Disillusionment” phase on their famous Hype Cyclearchitecture [2]. At the core of enterprise architecture [8].lies the idea that information systems and the However, on a deeper and more fundamentalbusiness should be considered holistically and level an eligible question is if the fundamentalconsistently, typically by the use of conceptual assumptions about enterprise architecture as an(architecture) models [3]. effective management support tool are correct. Many However, how and exactly for what purposes (by all means sound) arguments have been madeenterprise architecture is used varies greatly. Some often implicitly to support this assumption, but feworganizations use it for information technology (IT) empirical studies on this issue can be found.concerns alone, others for both business and IT. This article reports on results from a large surveySome focus on achieving a holistic and unified conducted among Swedish practitioners involved ingovernance approach and others care more for the enterprise architecture management (EAM). Themodeling issues and try to achieve consistent and all survey investigated the current scope and maturity as 1530-1605/11 $26.00 © 2011 IEEE 1
  2. 2. Proceedings of the 44th Hawaii International Conference on System Sciences - 2011well as experienced benefits of enterprise architecture The result presented in Ross et al. [1] indicatedmanagement. that there is a relationship between a mature The remainder of this paper is structured as enterprise architecture and business benefits.follows. Section two discusses related works in the Enterprise architecture management is seen as thefield of enterprise architecture and in the impact of path to a mature enterprise architecture. For thisenterprise architecture management. Section three purpose, several frameworks have been developed topresents the research methodology and the evaluate the maturity of enterprise architecturehypotheses tested. Section four presents the result. In management. Including [14], [17] ,[18] ,[19], [20],Section five this result is discussed and in section six and [21].conclusions are drawn. In contrast to the maturity stages defined in [1] these maturity models describe how to measure and2. Related works improve the maturity of enterprise architecture management activities, and not the architectures The practice and theory of enterprise architecture themselves.management is well described in literature. This Naturally, there is a significant overlap betweenincludes textbooks (e.g. [9] and [10]), conference and the elements discussed in different maturity models.workshop proceedings (e.g. HICSS. EDOC and For example, all ([14], [17], [18], [19], [20], andTEAR), journal articles (e.g. [11] and [12], [21]) state that the existence of an architectureframeworks (e.g. TOGAF [6] and Modaf [13]) and function within the organization is important and thattechnical reports (e.g. [14]). Literature often provides key architecture roles should be defined. To monitora rationale for initiating enterprise architecture efforts conformance to the architecture standards, constantlyand explains how to make them successful. evaluated enterprise architecture management There are several descriptions of the benefits processes and management of the quality of thesegained from EA practices. The first chapter in [9] processes are also discussed in all these maturitystates 13 business benefits of enterprise architecture models. Furthermore, they all stress the need toplanning and 12 benefits of planned IT systems. A involve, and communicate with, key stakeholdersgood enterprise architecture will according to [6] and within the organization. The use of an agreed[15] provide a more efficient IT operation, lead to reference method for enterprise architecturebetter return on existing investments, reduce risks management leads to maturity in [18], [19], and [20].associated with future investments and make In [18], [19], and [14] well defined enterpriseprocurement faster, simpler and cheaper. According architecture management processes are associatedto TOGAF [6] an effective enterprise architecture is with maturity and in [14] a more process basedan indispensible mean to achieve competitive enterprise architecture management is associated withadvantage through information technology. In fact, higher maturity than management seen as a project.TOGAF states that it is critical to business survival These enterprise architecture managementand success. Further, according to Zachman [16] maturity models describe the process of how goodsystems designed according to an EA are relatively enterprise architectures should be developed. Goodeconomical, quickly implemented, and easier to enterprise architecture is in turn also attributed todesign and manage. several benefits. This includes faster procurements, Evidence supporting the benefits from good increased return on investments and more efficient ITenterprise architecture management are however operations. This theory is found in both academic andsparse. In fact, none of the abovementioned texts practitioner-oriented literature. Yet, there are fewpresents evidence that support the claimed benefits. empirical studies that confirm the theory that mature Ross et al. [1] describes a model where enterprise enterprise architecture management leads to thesearchitecture maturity classified into four stages: benefits.business silos, standardized technology, optimized Slot et al. [22] have tested if well-defined solutioncore and business modularity. Advancing in these architectures impact project success variables. Theystages will lead to reduced IT costs, increased IT find that a solution architecture positively influence:responsiveness, improved risk management, schedule delays, budget overrun, variance in projectincreased management satisfaction and increased time frame and budget overrun, if the project deliversstrategic business impact [1]. These relationships are as expected and the technical fit. The definition of aalso supported by data. However, no peer-reviewed target architecture, or a to-be architecture, is apublication can be found where these results are fundamental part of enterprise architecturepresented. management (see for example [6]). The study described in [22] is however made in the context of 2
  3. 3. Proceedings of the 44th Hawaii International Conference on System Sciences - 2011software development projects and no similar study The rationale for these hypotheses and theirhas been found that generalizes these findings to an operationalization is described below.enterprise perspective. Boh and Yellin [11] present a comprehensive 3.1. IT-successstudy treating the relationship between enterprisearchitecture governance and the resulting enterprise No academic research papers on enterprisearchitecture. Based on a survey responded by 90 architecture have been found that explicitly describeorganizations they test if governance mechanisms for tests for the success of enterprise architectureenterprise architecture impacts the use and activities. Nor are such tests proposed in the booksconformance to enterprise architecture standards. and standards available.They also test if conformance to enterprisearchitecture standards impact four outcomes in the Table 1. Hypotheses tested.enterprise architecture: the heterogeneity in the IT IT-successinfrastructure, the replication of services in the IT departments’ satisfied Successful Execution procurement projectsinfrastructure, business application integration and of IT-projectsenterprise data integration. Their result confirms the Operational Duration ofhypothesis that governance mechanisms have a with ITbeneficial impact on the use of enterprise architecturestandards and the four outcomes. As Boh and Yellin [11], this paper presents a testof enterprise architecture management’s impact onorganizations. Unlike Boh and Yellin, however, thispaper tests outcomes in terms of direct business Existence of H1.1 H1.2 H1.3benefits. Namely the percent of IT projects that are EAMsuccessful, the duration of IT procurement projects, EAM Activitiesand how satisfied the business is with the Amount oforganizations information technology. time worked H2.1 H2.2 H2.3 Further, unlike Boh and Yellin this paper tests the with EAMoutcomes against the maturity of the enterprisearchitecture management function and the duration Maturity offor which this function has existed. As described H3.1 H3.2 H3.3above the maturity models developed for enterprise EAMarchitecture management activities describe a largenumber of variables that indicate maturity. In this As described in chapter 2 and parts of the mainpaper enterprise architecture maturity is measured literature in the enterprise architecture field one canthrough variables that are present in several maturity implicitly state testable hypotheses. These are statedframeworks. The measurement instrument (survey) as benefits of enterprise architecture either for theused in this study is an adapted version of the one business directly or via the impact of improved IT-used in [23][24][25]. However, none of the previous management.studies with this survey instrument test the impact ofenterprise architecture management of IT success in 3.1.1. Successful execution of IT-projects.enterprises. Enterprise architecture management is believed to reduce the amount of IT-projects that fail within the3. Research methodology organizations. For instance, Lankhorst [15] state that EA leads to “Reduced risk overall in new investment, The overall aim of the presented study is to test and the cost of IT ownership” and to “Easier upgradethe general hypothesis that Enterprise Architecture and exchange of system components”. The OpenManagement activities has a positive impact on IT- Group [6] also states that EA leads to reduced risk forsuccess. future investments. To assess this impact three main hypothesis The variable “Successful execution if IT-projects”evaluated through nine sub hypotheses are tested. was used to assess the level of IT-project risks. ThisThese are, as outlined in Table 1, derived from three variable is a composite of two items. The firstvariables of IT-success and three variables of addresses the average budget overrun when investingenterprise architecture management activities. in new information technology. The second one covers the average (schedule) delay of information 3
  4. 4. Proceedings of the 44th Hawaii International Conference on System Sciences - 2011technology projects. Both are determined relative to 3.2.1. Existence of enterprise architectureother organizations in the same field and of the same management. In our survey we used the age of thesize. organizational function for Enterprise Architecture Management to determine whether EAM exists or3.1.2. Duration of procurements projects. Mature not. The participants had the possibility to state theenterprise architectures and thus mature enterprise age of their EAM on the scale {does not exist, <1, 1-architecture management has been claimed to 2, 2-5, >5 years}. All positive values were treated asdecrease the duration of IT procurement projects. indicators of the existence of EAM.TOGAF [6] does for instance state that enterprisearchitecture makes the procurement process faster 3.2.2. The amount of time working with enterprisewithout sacrificing architectural coherence. TOGAF architecture management. The amount of timealso state that risks (e.g. due to schedule overruns) worked with EAM was evaluated through the sameare decreased with good enterprise architecture question as the one used to evaluate the existence. Inmanagement. Zachman [16] describes the descriptive contrast to a binary answer, we considered the five“as-is” architecture as a necessary condition for different time intervals to gather the actual time ofchanging an enterprise safely. the EAM duration. The variable Duration of procurement projectswas built upon the comparison of the duration for 3.2.3. Enterprise architecture managementprocurement to organizations of the same size and maturity. As described in section 2 there are severaloperating within the same field. One item was used to maturity models for enterprise architecturedetermine this variable. management. As the union of these produces a large number of variables, which are partly overlapping, a3.1.3. Operational departments’ satisfied with IT. subset of the maturity-variables was selected.The third variable associated with IT-success Seven different items (questions) was used toincluded in this study was the “Operational assess the maturity level of enterprise architecturedepartments’ satisfied with IT.” activities (cf. Table 2). These items are drawn from The basic notion behind enterprise architecture the following maturity frameworks: [14], [17], [18],activities is to create a value for the enterprise. [19], [20], and [21].Enterprise architecture is intended to increase the The arithmetic mean of these seven items wasreturn on IT-investments, reduce risks associated used. As can be seen from Table 2 the items arewith IT-projects, make procurements simpler and included in the majority of the maturity frameworksfaster, and much more. The operational departments discussed earlier. For instance, all do for instancewithin an enterprise are typically those that benefit discuss the role of architects. Three out of six stressfrom such qualities. It can also be expected that that enterprise architecture should be seen as process,operational departments will have to bear the costs of hence that budgets should be provided continuously.enterprise architecture management activities, either Four out of six maturity models associate an agreeddirectly or indirectly. If enterprise architecture reference method to mature enterprise architecturemanagement activities are beneficial it could management practices.therefore be expected that these make operationaldepartments more satisfied with the IT. Table 2. Operationalization of EAM maturity. This variable is defined as the satisfaction Items Fromexperienced by the organizations operational The roles of the architects are [14][17][18]department when compared to the satisfaction of defined. [19][20] [21]other organizations’ operational departments. Budget and resources of EAM are [14][18][20] 3.2. Enterprise architecture management provided continuously activities There exist defined EAM-processes. [14][18][20] Three variables were defined to assess the (Enterprise-)Architecture models are [14][17][18]enterprise architecture management activities within evaluated and benchmarked [19][20]the organization. The existence and age of the constantly.enterprise architecture management function within EAM-processes are evaluated and [14][17][18]the organization are two of these. The maturity of benchmarked constantly. [19][20]enterprise architecture management is the third. 4
  5. 5. Proceedings of the 44th Hawaii International Conference on System Sciences - 2011There exist defined maintenance [14][18] • H3.2 - The maturity of EAM correlates withprocesses for (Enterprise-) [19][20] duration of procurement projects.Architecture models and data. • H3.3 – The maturity of EAM correlates withEAM is performed with use of an [17][18] the satisfaction of operational departments.agreed (reference-) method. [19][20] 4. Result 3.3. Hypotheses The overall aim of the presented study is to test three main hypotheses (divided into nine sub- With the above as a basis, three main hypotheses hypotheses) related to Enterprise Architectureon Enterprise Architecture Management’s effects on Management and its effects on the organizationsIT-success were tested proposed: performance in IT projects and IT products. • H1: The existence of Enterprise Architecture Management function correlates with IT- 4.1. Questionnaire and data collection success. • H2: The amount of time the organization has A survey questionnaire was prepared. The greater worked with Enterprise Architecture part of the questions in the survey has been used in Management correlates with IT-success. several research projects at the University of St • H3: The maturity of Enterprise Architecture Gallen in Switzerland, thus also tested and iteratively Management correlates with IT-success. improved [23]. The additions concern the questions These three main hypotheses are detailed into regarding the execution of IT-projects and the IT-three hypotheses each to consider the three types of satisfaction in operational departments.IT-success described above: the number of The survey was distributed during an Enterprisesuccessfully executed IT-projects, the duration of Architecture Symposium at the Royal Institute ofprocurement projects and the satisfaction of Technology in Stockholm, Sweden the 15th ofoperational departments. October 2009. The symposium had 268 attendants The first hypothesis describing the relationship mainly from Swedish companies. Of the 268between the existence of EAM and IT-success is attendants 68 handed in surveys.tested through an analysis of the Hypothesis H1.1 to As the population used for this survey is personsH1.3. The three sub-hypotheses in total capture the attending an event there is a risk that it is biased. Ininitial hypothesis H1, but allow a more detailed particular, it can be expected to be skewed towardsanalysis. The same goes for Hypothesis H2 and H3 organizations with an interest in enterprisewhich are detailed in H2.1 to H2.3 and H3.1 to H3.3 architecture management. Most of the respondentsrespectively. work in large organizations – 73 percent work in The in total nine hypotheses are: organizations with more than thousand employees. • H1.1 – The existence of EAM correlates with Most industry areas are covered. However, as many the share of successfully executed IT- as 69 % works in an IT-department, cf. Table 3, projects. Table 4 and Table 5. • H1.2 – The existence of EAM correlates with the duration of procurement projects. Table 3. Survey participants by business. • H1.3 – The existence of EAM correlates with Industry Count Percentage the satisfaction of operational departments. • H2.1 – The amount of time worked with Banking 13 19% EAM correlates with share of successfully Commerce 2 3% executed IT-projects. Consulting 5 7% • H2.2 – The amount of time worked with EAM correlates with duration of procurement Energy 3 4% projects. Insurance 3 4% • H2.3 – The amount of time worked with Manufacturing 10 15% EAM correlates with the satisfaction of operational departments. Public Administration 13 19% • H3.1 – The maturity of EAM correlates with Software Vendor/Developer 3 4% the share of successfully executed IT- Telecommunication 4 6% projects. Other 12 18% 5
  6. 6. Proceedings of the 44th Hawaii International Conference on System Sciences - 2011 Table 4. Survey participants by enterprise histograms quantile-quantile plots for all studied size. variables. These tests showed normality.Size of company Count Percentage As the questionnaire used for this study is based on questionnaire that has been used and tested20–99 0 0% multiple times in earlier studies the reliability of the100–249 5 8% origin is to a certain extent transferable. However,250–499 6 10% three elements were added to assess the output produced in terms of information system500–1000 6 10% deliverables. Two of these were combined into the> 1000 19 30% variable Successful Execution of IT-projects and> 10 000 26 42% seven of the original elements were combined into the variable survey Enterprise Architecture Management Maturity. The reliability of these two items was measured with the Cronbach’s Alpha reliability coefficient, originally proposed in 1951 Table 5. Background of the enterprises’ [27] and revised and discussed in [28]. Table 6 respondents. shows the result of this test.Area of activity Count Percentage Table 6 Reliability assessment usingIT-department 46 69% Cronbach’s Alpha.Management 6 9% Variable Items α The role of the architects isOperational dep. 8 12% defined.Other 7 10% Budget and resources of EAM are provided continuously There exist defined EAM- processes. The demographical differences between the (Enterprise-)Architecturesurvey respondents were not considered in the models are evaluated andanalysis. However, this could be an interesting follow EAM benchmarked constantly. 0.881up analysis, since there might be significant trends to Maturity EAM-processes are evaluatedbe found in these differences. and benchmarked constantly. The survey consisted of three parts. Part one There exist definedcontained questions regarding the background of the maintenance processes forrespondents such as industry and company. The (Enterprise) Architecturesecond part contained five questions regarding IT- models and data.investments and IT-procurement. Part three of the EAM is performed with usesurvey had five sections. These five sections of an agreed (reference)addressed the Introduction, Understanding, method.Organization, Functions and Planning processes of Compared to otherEnterprise Architecture Management. For each organizations within the samestatement the respondents were asked to mark the sector and of the same size,current and target situation on a five-point Likert my organization has a lowscale (where 1 represent strongly disagree, 2 average budget overrun whendisagree, 3 partly agree, 4 agree and 5 equals strongly investing in new informationagree). Successful technology. Execution 0.717 Compared to other of IT 4.2. Validity and reliability organizations within the same sector and of the same size, The variables chosen for testing the hypotheses my organization has a loware described in section 3. This data set was screened average (schedule) delay ofaccording to the checklist provided by [26], e.g. information technologyproofreading worksheet against original data sources, projects.checking for outliers, missing values and normality.The test for normality was implemented by printing The threshold of the Alpha coefficient is dependent on the number of items being assessed. 6
  7. 7. Proceedings of the 44th Hawaii International Conference on System Sciences - 2011Interpreting the results of Cortina [29], seven items As can be seen in Table 7 the number ofwith Alpha 0.881 and two items at 0.717 gives an responses varies between 42 and 57 for the differentacceptable reliability level. hypotheses. There majority of these correlations are As described in chapter 2, there is no consensus both weak and insignificant. The hypotheses withon how to measure the maturity of enterprise strongest correlation are those related to hypothesis 3architecture management, i.e. what items Enterprise – the influence of mature enterprise architectureArchitecture Management Maturity should comprise management on IT success. The two strongestof. As no standard has been commonly accepted it correlations, H3.2 and H3.3, are also significant todifficult to test content. As outlined in section 3 the the 0.05-level. Hypothesis H3.1 is almost significantitems included in this survey are all well represented to the 0.10-level.in frameworks developed to measure this variable.While they do not cover all the elements that 5. Discussion and research outlookdifferent frameworks find relevant they capture anumber of essential variables. This section is divided into three parts: discussion The lack for commonly accepted maturity- of results, industrial and academic impact and futuremeasurement standards makes it difficult to test work.convergent validity for the used EAM maturityvariable. Table 7. Pearson correlation for each To test disciminant validity one would like to hypothesis.ensure that the used measure for enterprise Hypothesis Pearson Correlation Sig. (2-tailed) Narchitecture management maturity is different frommeasures of variables that are similar to enterprise H1.1 0.092_ 0.506 55architecture management maturity but represent some H1.2 0.074_ 0.633 44other construct. IT governance is a discipline closely H1.3 -0.088_ 0.513 57related to enterprise architecture management and atest for disciminant validity could ensure that the H2.1 -0.090_ 0.563 44operationalization used in this study is different from H2.2 -0.121_ 0.441 43IT governance maturity. However, as no commonlyaccepted standards for IT governance maturity exists H2.3 .a_ 0.000 42either, such tests are difficult to construct. However, H3.1 0.238_ 0.104 48the items of EAM maturity focus on architects and H3.2 0.304* 0.035 48enterprise architecture management processes. Asthese aspects are not the locus of IT governance H3.3 0.292* 0.044 48frameworks it appears as unlikely that disciminant .a Cannot be computed because at least one of thevalidity is threatened. Also, the context under which variables is constant.the respondents filled in the survey (an enterprisearchitecture symposium) make reasonable to believe * Correlation is significant at the 0.05 level (2-tailed).that the respondents provided answers that reflect N Represents the number of respondents answeringtheir work within enterprise architecture questions used to test this hypothesis.management. The operationalization of successful executed 5.1. Results discussionprojects used in this study is a composite of projectsbeing on time and within budget. According to the Our study indicates that the mere existence of anmost common definition of successful projects the enterprise architecture function at a company is notquality delivered should also be included in this enough if the goal is to have IT-success. The resultsoperationalization. There are also many alternative found also indicate that having an enterprisedefinitions of project success [30]. In relation to these architecture function for a long time will not impactthe operationalization used in this study only cover a the success of IT. However, since the correlations arepart of the construct “Successful IT project”. insignificant these two results do not mean that the two related hypotheses tested can be rejected. 4.3. Test result Dividing the three main hypotheses our study shows that the enterprise architecture management The hypotheses were tested with standard maturity correlates with all three IT-successbivariate two-tailed Pearson correlation. The results measures, namely successful execution of IT-are presented in Table 7. projects, duration of procurement projects and 7
  8. 8. Proceedings of the 44th Hawaii International Conference on System Sciences - 2011operational departments’ satisfied with IT, whereas with slightly different maturity definitions to test ifthe last two of these are signification at the 0.05 the correlation still exists.level. Further, the IT-project success definition might be A common question to ask when performing considered by some to be incomplete. For instance,statistical studies like this is if the correlations found some might argue that project quality is notalso indicate causality. It appears unlikely that the addressed sufficient enough. It is partly covered bycorrelation is due to an inverse causality, e.g. that the subjective measure of operational departmentbusiness satisfaction caused the enterprise happiness. Still, this can also be addressed in a futurearchitecture management maturity. No theory studied study. Moreover the metric used here is relative. Itdid either suggest this. It is of course possible that might still be the case the project success is quitesome other (third, unmeasured) factor causes the poor in absolute terms.found correlation. However, it is not easy to come up In this study company representatives werewith such a relationship that at the same time surveyed during a symposium focusing on enterprisecorroborates H3 and rejects H1 and H2. For example, architecture. It is our firm belief that thethat the organization is “rich” could perhaps explain a organizations at the symposium provide a goodsatisfied business (because they are not under stress enough sample giving survey answers whichin general and that there is enough room for having indicates the impact of enterprise architecturetime to develop a mature enterprise architecture management on IT-success. However, in order to getmanagement), but with these arguments there is no a more representative population for the analysis andexplanation of the lack of correlation with the a larger sample of enterprises the study could be doneexistence of enterprise architecture management. in another way, for example by sending the survey to However, given that the correlation was not very a random set of organizations.strong, it is apparent that IT-project success and In addition to corroborate the findings presentedbusiness satisfaction also depends on other factors here, interesting further work is to dig deeper into thethan enterprise architecture management maturity. concept of maturity and investigate if there exists any specific aspect(s) of “being mature” that impacts the 5.2. Industrial and academic impact success of running IT satisfaction of the business. How important is for instance the use of models for For the practitioner this study indicates that when enterprise architecture management? Is it more orworking with enterprise architecture it is important to less important than for instance the enterprisebe active and aim at a mature enterprise architecture architecture management processes? In other words,management function. Time itself will not do the are there any success factors for enterprisetrick. The definition used in this study of EAM can architecture management? Moreover, with a betterguide a practitioner in EAM maturity work. understanding of the importance of “smaller units” of In academia enterprise architecture is getting maturity it might also be possible to build up anmore and more research focus. This study shows that empirically based enterprise architectureresearchers working with enterprise architecture management maturity model. If it turns out thatmanagement are doing important research for models for instance are of no help unless there is aindustry, since enterprise architecture management process in place, this should be reflected in a maturitycan help companies become better at IT. This study index in order to make it really useful for thecan also help researchers to focus their studies. For practitioners.instance, since enterprise architecture managementmaturity shows significant correlation with IT- 6. Conclusionssuccess researchers can help practitioners by studyinghow the maturity items can be implemented This paper provides an empirical argument for thesuccessfully. hypothesis that enterprise architecture management needs to be mature in order to have effect. On the 5.3. Future work other hand, the mere existence of enterprise architecture work has not been found to have an The definition chosen for enterprise architecture impact. Nor does the amount of time the enterprisemanagement maturity in this study consists of seven architecture function has existed appear to impact theitems all widely considered in enterprise architecture success of IT within enterprises.research literature. However, others might use a These conclusions are based on data collecteddifferent definition than the one employed here. during an event in Stockholm 2009. Future workThus, a future study could test the same hypotheses could the findings in this study with a more 8
  9. 9. Proceedings of the 44th Hawaii International Conference on System Sciences - 2011representative and encompassing population of [15] TOGAF, ArchiMate Technicalenterprises. Further work could also be made on the Specification, van Haren Publishing, 2009.development and of validation measurement [16] J.A. Zachman, "Enterprise architecture: Theinstruments for measuring enterprise architecture issue of the century," Databasemanagement activities and their impact. Programming and Design, vol. 10, 1997, pp. 44-55.7. References [17] US Goverment Accountability Office, "A Framework for Assessing and Improving[1] J.W. Ross, P. Weill, and D.C. Robertson, Enterprise Architecture Management," 2003. Enterprise architecture as strategy: Creating [18] M.E. van Steenbergen, "Architecture a foundation for business execution, Maturity Matrix - DYA," 2005. Massachusetts, USA: Harvard Business [19] M. van Steenbergen, M. van Den Berg, and School Press, 2006. S. Brinkkemper, "An instrument for the[2] S. Kurpjuweit and R. Winter, "Viewpoint- development of the enterprise architecture based meta model engineering," EMISA practice," ICEIS, Funchal, Madeira, 2007, 2007. Portugal: 2007, pp. 12-16.[3] R. Winter and R. Fischer, "Essential Layers, [20] J. Schekkerman, "Extended Enterprise Artifacts, and Dependencies of Enterprise Architecture Maturity Model Support Guide Architecture," Journal of Enterprise v2. 0," 2006. Architecture, vol. 3, 2007, p. 7–18. [21] NASCIO, "NASCIO Enterprise[4] H. Koning, R. Bos, and S. Brinkkemper, "A Architecture Maturity Model, v 1.3," 2003. Lightweight Method for the Modeling of [22] R. Slot, G. Dedene, and R. Maes, "Business Enterprise Architectures: Introduction and Value of Solution Architecture," E. Proper, Empirical Validation," Utrecht: Department F. Harmsen, and J.L. Dietz, of Information and, 2006. Berlin/Heidelberg, Germany: Springer[5] F. Theuerkorn, Lightweight enterprise Verlag, 2009, pp. 84-108. architectures, CRC Press, 2004. [23] C. Riege and S. Aier, "A Contingency[6] TOGAF, The Open Group Architecture Approach to Enterprise Architecture Method Framework (TOGAF) - version 9, The Open Engineering," Pre-Proceedings of the 3rd Group, 2009. Workshop on Trends in Enterprise[7] D. Hinchcliffe, "Pragmatic new models for Architecture Research (TEAR 2008), Dec 1, enterprise architecture take shape," 2009. 2008, in conjunction with the 6th[8] Gartner, "Gartner Hype Cycle for Enterprise International Conference on Service Architecture 2010," 2010. Oriented Computing (ICSOC08), Sydney,[9] S.H. Spewak, Enterprise Architecture Australia: Springer, 2009, pp. 388-399. Planning, Princeton, NJ, USA,: John Wiley [24] S. Aier, C. Riege, and R. Winter, and Sons, inc., 1992. "Classification of Enterprise Architecture[10] M. Lankhorst, Enterprise architecture at Scenarios–An Exploratory Analysis," work: Modelling, communication and Enterprise Modelling and Information analysis, Heidelberg, Berlin, Germany: Systems Architectures, vol. 3, 2008, p. 14– Springer-Verlag, 2009. 23.[11] W. Boh and D. Yellin, "Using Enterprise [25] S. Aier, C. Riege, and R. Winter, Architecture Standards in Managing "Unternehmensarchitektur – Information Technology," Journal of Literaturüberblick und Stand der Praxis," Management Information Systems, vol. 23, Wirtschaftsinformatik, vol. 50, 2008, pp. 2007, pp. 163-207. 292-304.[12] R. Lagerström, "Enterprise Systems [26] R. Warner, Applied Statistics – From Modifiability Analysis - An Enterprise Bivariate Through Multivariate Techniques, Architecture Modeling Approach for London: SAGE Publications Inc., 2008. Decision Making," Architecture, 2010. [27] L.J. Cronbach, "Coefficient alpha and the[13] MoD, "MODAF Handbook, technical internal structure of tests," Psychometrika, specification for MODAF," 2005. vol. 16, 1951, pp. 297-334.[14] DoC, "Enterprise Architecture Program [28] L.J. Cronbach and R.J. Shavelson, "My Support," 2007. Current Thoughts on Coefficient Alpha and Successor Procedures," Educational and 9
  10. 10. Proceedings of the 44th Hawaii International Conference on System Sciences - 2011 Psychological Measurement, vol. 64, 2004, pp. 391-418.[29] J.M. Cortina, "What is coefficient alpha? An examination of theory and applications.," Journal of Applied Psychology, vol. 78, 1993, pp. 98-104.[30] R. Atkinson, "Project management: cost, time and quality, two best guesses and a phenomenon, its time to accept other success criteria," International Journal of Project Management, vol. 17, 1999, pp. 337-342. 10