Cloud enabled business process management systems

• 1.City University London MSc in Business Systems Analysis and Design Project Report2011Cloud Based Business ProcessManagement SystemsJafar Railton Supervised byBill Karakostas 23 September 2011

2. By submitting this work, I declare that this work is entirely my ownexcept those parts duly identied and referenced in my submission.It complies with any specied word limits and the requirements andregulations detailed in the coursework instructions and any otherrelevant programme and module documentation. In submittingthis work I acknowledge that I have read and understood the reg-ulations and code regarding academic misconduct, including thatrelating to plagiarism, as specied in the Programme Handbook.I also acknowledge that this work will be subject to a variety ofchecks for academic misconduct.Signed: 3. Acknowledgements In the Name of Allah, the Most Benecent, the Most Merciful All praise and thanks are due to Allah and may peace andblessings be upon the Messenger of AllahTo proceed:I would like to thank my family for their patience and support sosorely needed by this incorrigible student, Jack of many Masters.I would also like to thank my supervisor, Bill Karakostas, for hisguidance and support throughout, which greatly assisted in thesuccess of this endeavour. 4. AbstractAlthough both business process management (BPM) and cloudcomputing are relatively well-dened in the literature, their recentamalgamation in the form of Cloud BPM is not. This researchcontributes to the literature on Cloud BPM, rstly by dening itsterms, and then by considering its application and merits. Themethods employed are an exhaustive literature survey of the sub-ject domain, followed by the generation of a hypothesis regardingthe denition of Cloud BPM. An online survey questionnaire isused to test the hypothesis by collecting data from a target groupof BPM practitioners. The ndings will be of interest to potentialconsumers of cloud based BPM systems, as well as to vendors ofBPM systems, and analysts seeking to advise on the potential onthis emerging technology and how it might help customers realizetheir business goals.Keywords: cloud computing, business process management, cloudbased BPM, BPMS 5. ContentsContentsivList of Figures viiList of Tablesix1 Introduction and project objectives1 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 1.2 Problem statement . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.3 Important note on terminology . . . . . . . . . . . . . . . . . . 2 1.4 Aims and objectives . . . . . . . . . . . . . . . . . . . . . . . . 2 1.5 Applicable methodologies . . . . . . . . . . . . . . . . . . . . .3 1.6 Project contribution and supposed beneciaries . . . . . . . . .4 1.7 Organization of this project report . . . . . . . . . . . . . . . . 42 Literature survey6 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 2.2 Cloud BPM from the linguistic point of view . . . . . . . . . . 6 2.3 Business process management . . . . . . . . . . . . . . . . . . . 7 2.3.1 Dening BPM . . . . . . . . . . . . . . . . . . . . . . . 7 2.3.2 The BPM lifecycle . . . . . . . . . . . . . . . . . . . . . 8 2.3.3 The BPM discipline . . . . . . . . . . . . . . . . . . . .9 2.3.4 BPM technology: the BPM suite . . . . . . . . . . . . . 9 2.3.5 BPM and service oriented architecture . . . . . . . . . .12 2.3.6 BPM adoption and potential obstacles to be overcome .14 2.4 Cloud computing . . . . . . . . . . . . . . . . . . . . . . . . . .16 2.4.1 Dening cloud computing . . . . . . . . . . . . . . . . .17 2.4.2 Characteristics of cloud computing . . . . . . . . . . . . 19iv 6. CONTENTS2.5 Cloud BPM . . . . . . . . . . . . . . . . . . . . . . . . . . . . .212.5.1 History . . . . . . . . . . . . . . . . . . . . . . . . . . .212.5.2 Vendor oerings . . . . . . . . . . . . . . . . . . . . . .222.5.3 Analyst point of view . . . . . . . . . . . . . . . . . . .292.5.4 Practitioners, bloggers, commentators . . . . . . . . . .302.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .323 Methods333.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333.2 Literature survey . . . . . . . . . . . . . . . . . . . . . . . . . .343.2.1 Literature search . . . . . . . . . . . . . . . . . . . . . .343.2.2 Literature review . . . . . . . . . . . . . . . . . . . . . .353.3 Online survey . . . . . . . . . . . . . . . . . . . . . . . . . . . .353.3.1 Survey design . . . . . . . . . . . . . . . . . . . . . . . .363.3.2 General considerations . . . . . . . . . . . . . . . . . . . 363.3.3 Survey target . . . . . . . . . . . . . . . . . . . . . . . .383.3.4 Motivation of questions . . . . . . . . . . . . . . . . . .383.4 Proposed denition . . . . . . . . . . . . . . . . . . . . . . . . . 453.5 Evaluation of proposed denition . . . . . . . . . . . . . . . . . 453.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .454 Results464.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464.2 Literature analysis . . . . . . . . . . . . . . . . . . . . . . . . .464.2.1 Description of Cloud BPM . . . . . . . . . . . . . . . . 464.2.2 Hypothesis statement . . . . . . . . . . . . . . . . . . . 484.3 Survey questionnaire results . . . . . . . . . . . . . . . . . . . . 494.3.1 Respondent-specic information . . . . . . . . . . . . . 494.3.2 Dening Cloud BPM . . . . . . . . . . . . . . . . . . . .514.3.3 Characterizing Cloud BPM . . . . . . . . . . . . . . . . 544.3.4 Cloud BPM - pros and cons . . . . . . . . . . . . . . . .584.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .595 Discussion 615.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615.2 Discussion of online survey results . . . . . . . . . . . . . . . .615.2.1 Respondents . . . . . . . . . . . . . . . . . . . . . . . .61v 7. CONTENTS5.2.2 Dening Cloud BPM . . . . . . . . . . . . . . . . . . . . 625.2.3 Characterizing Cloud BPM . . . . . . . . . . . . . . . .635.2.4 Cloud BPM - pros and cons . . . . . . . . . . . . . . . . 675.3 Status of the hypothesis . . . . . . . . . . . . . . . . . . . . . .675.4 Implications of these ndings for the future of Cloud BPM . . . 686 Evaluation, Reections, Conclusions 716.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . .716.2 Summary of project . . . . . . . . . . . . . . . . . . . . . . . .716.3 Evaluation of methods used and outcomes achieved . . . . . . .726.4 Suggestions for further research . . . . . . . . . . . . . . . . . .736.5 Some personal reections . . . . . . . . . . . . . . . . . . . . .73References75A Project denition for MSc in Business Systems Analysis andDesign A-1B Survey TargetB-1C Online SurveyC-1D Summary of results for Likert scale questionsD-1E BPM Twitter list E-1F BPM Findings F-1G Online survey results summaryG-1 vi 8. List of Figures 1.1 Organization of Cloud BPM project . . . . . . . . . . . . . . .3 2.1 The BPM lifecycle . . . . . . . . . . . . . . . . . . . . . . . . .9 2.2 Components of a BPMS . . . . . . . . . . . . . . . . . . . . . .10 2.3 Spectrum of business processes . . . . . . . . . . . . . . . . . .13 2.4 BPM and SOA . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 2.5 Cloud computing . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.6 Bonita Studio . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.7 Cordys cloud platform . . . . . . . . . . . . . . . . . . . . . . . 25 2.8 Cordys Business Operations Platform . . . . . . . . . . . . . . 26 2.9 Intalio|BPM Architecture . . . . . . . . . . . . . . . . . . . . .28 3.1 Cloud BPM project process . . . . . . . . . . . . . . . . . . . . 33 4.1 Survey respondents by BPM role . . . . . . . . . . . . . . . . .50 4.2 Survey respondents by company size . . . . . . . . . . . . . . .51 4.3 Survey respondents by company sector . . . . . . . . . . . . . .52 4.4 Survey results cloud BPM functionality . . . . . . . . . . . .53 4.5 The primary advantages of Cloud BPM . . . . . . . . . . . . . 58 4.6 Stated advantages of Cloud BPM . . . . . . . . . . . . . . . . .60 4.7 Stated disadvantages of Cloud BPM . . . . . . . . . . . . . . . 60 5.1 Cloud ecosystem with BPM as hub . . . . . . . . . . . . . . . . 70 A.1 Schedule of work - Gantt chant . . . . . . . . . . . . . . . . . . A-5 B.1 Post to LinkedIn BPM groups . . . . . . . . . . . . . . . . . . . B-2 C.1 Cloud BPM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-1vii 9. LIST OF FIGURESC.2 Dening Cloud BPM . . . . . . . . . . . . . . . . . . . . . . . . C-2C.3 Characterizing Cloud BPM (1) . . . . . . . . . . . . . . . . . . C-3C.4 Characterizing Cloud BPM (2) . . . . . . . . . . . . . . . . . . C-4C.5 Cloud BPM - Pros and Cons . . . . . . . . . . . . . . . . . . . C-5C.6 About You . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-6C.7 The End . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-7 viii 10. List of Tables 4.1 Survey results Q1.1 . . . . . . . . . . . . . . . . . . . . . . . . 52 4.2 Survey results Q1.3 . . . . . . . . . . . . . . . . . . . . . . . . 53 4.3 Survey results Q2.1 . . . . . . . . . . . . . . . . . . . . . . . . 54 4.4 Survey results Q2.2 . . . . . . . . . . . . . . . . . . . . . . . . 55 4.5 Survey results Q2.3 . . . . . . . . . . . . . . . . . . . . . . . . 55 4.6 Survey results Q2.4 . . . . . . . . . . . . . . . . . . . . . . . . 55 4.7 Survey results Q2.5 . . . . . . . . . . . . . . . . . . . . . . . . 56 4.8 Survey results Q2.6 . . . . . . . . . . . . . . . . . . . . . . . . 56 4.9 Survey results Q2.7 . . . . . . . . . . . . . . . . . . . . . . . . 57 4.10 Survey results Q2.8 . . . . . . . . . . . . . . . . . . . . . . . .57ix 11. 1. Introduction and projectobjectives1.1IntroductionBusiness process management (BPM) is a mature business discipline that hasspawned a number of technologies to support it. Businesses now put pro-cess rst (Ould, 2005, p2), and BPM technologies have evolved to supporta business user centred approach to BPM. These days, organizations dependon BPM to enable them to adapt to rapidly changing business conditions byenabling the design and execution of business processes that can span the en-tire enterprise, as well as connect with systems external to it. Today it is theagile who survive those organizations who are able to adapt to change, toinnovate as well as continuously improve, and to continuously monitor andanalyze the results of these adaptations. In the current web enabled business environment, processes in many casesdepend on the discovery and recognition of components that exist as webservices (Datamonitor, 2009) and BPM systems must facilitate this. Fur-thermore, the current trend is towards increased emphasis on mobility andcollaboration as essential elements to support the agility and currency of busi-ness processes. This means that BPM vendors are increasingly seeking toaugment their BPM packages by incorporating greater Web 2.0 type function-ality. Cloud based BPM is one response to these new demands.1.2Problem statementAlthough cloud based BPM is already a reality, it is an emerging technologyand still evolving; thus, it is not entirely clear what it is that BPM vendors 1 12. 1. INTRODUCTIONare oering in the cloud, and how cloud based BPM can be evaluated as avalue-adding business proposition.1.3 Important note on terminologyAs will be seen later in the discussion, the concept of the provision of BPMtools and systems using cloud computing technologies is referred to in vari-ous terms, such as cloud based BPM, cloud enabled BPM, BPM in thecloud, BPM as a service, BPM on demand, etc. This project proposes toinvestigate the commonality of these terms under the umbrella concept CloudBPM.1.4 Aims and objectivesThe objective of the project is to propose a denition of the concept CloudBPM. The validity of the proposed denition will rest upon an evaluation ofits utility in further clarifying the key issues of the problem domain. In order to arrive at such a denition, the project rst presents a reviewof the literature surrounding Cloud BPM, and analyses what vendors andanalysts intend through references to cloud based BPM, cloud enabledBPM, BPM in the cloud, BPM as a service, BPM on demand, andso on. This analysis focuses on two key aspects of Cloud BPM technology:functionality and architecture. Further objectives of this project are to identify: (1) any dierentiatingfeatures of Cloud BPM over on-premise BPM, (2) the advantages and disad-vantages of Cloud BPM, and (3) future trends relating to Cloud BPM. Someother questions that inform the research are: What types of BPM software are available on the cloud? Does BPM on the cloud oer any particular advantages above and be-yond the advantages of SaaS considered generically? Are there any technical barriers to entry that apply to cloud based BPMwhich are not applicable to other types of SaaS oerings, e.g. CRM? Does BPM have any characteristics that make it particularly well-suitedto being deployed in the cloud?2 13. 1. INTRODUCTIONIn support of the above aims, the project also presents relevant backgroundinformation on BPM and cloud computing in general, as well as ancillarytechnologies such as service oriented architecture (SOA).1.5Applicable methodologiesThe research follows a sequential process (as described by Dawson (2009,p20)):1. Review the eld;2. Build a theory;3. Test the theory;4. Reect and integrate.This process, and how it is structured within this report, is represented inFigure 1.1 below.Figure 1.1: Organization of Cloud BPM project The foundation of the project consists of a literature review which cov-ers information from BPM vendors, analysts and commentators, as well back-ground information on BPM, cloud computing, and enabling architecture (e.g.SOA). From this body of information a tentative denition of Cloud BPM (the3 14. 1. INTRODUCTIONhypothesis) is proposed. The proposed denition of Cloud BPM is then testedagainst the opinion of experts in the eld, who were asked to complete a ques-tionnaire about BPM and its application in the cloud, designed specically forthat purpose.1.6 Project contribution and supposed beneciariesAlthough BPM is a mature discipline supported by similarly mature tech-nologies, the debate surrounding the denition of cloud computing continuesapace. As for Cloud BPM, the amalgam of BPM and cloud computing, it isvery much an emerging technology, with many vendors currently just begin-ning to enter the market while others remain on the sidelines still. Such beingthe case, Cloud BPM is not yet well-dened; consequently, the discussion re-garding the merits of its application lack rigour. This project will address thislack and thereby attempt to stimulate and further the discussion by proposinga research based denition of this important, emerging technology, which hasthe potential to disrupt the current BPM market certainly, and perhaps, theenterprise systems market in general as well. This research will therefore be of benet to businesses who are consideringimplementing BPM and considering buying BPM as a software or platform asa service. The research will also interest independent analysts and commen-tators, as well as other researchers in the eld of IT strategy.1.7 Organization of this project reportThis project report is divided into ve chapters. This initial chapter chapterhas introduced the project aims and objectives, and touched upon the meth-ods to be employed in meeting them. The next chapter (Chapter 2) consistsof a literature survey covering the key issues relating to the project domain.Chapter 3 goes on to describe the methods used to investigate the researchtopic, methods which included a literature review and a survey questionnaire.The results of these investigations are presented in Chapter 4 and a hypothesis a tentative denition of Cloud BPM is proposed. The project hypothe-sis is evaluated and further issues arising from the research are discussed inChapter 5. Finally, Chapter 6 concludes the project with a reection uponthe choice of project methods and their execution, followed by an evaluation 4 15. 1. INTRODUCTIONof project outcomes. A personal view of the projects benets and lessonslearned is also oered.5 16. 2. Literature survey2.1 IntroductionThis literature review presents an examination of the key concepts to be con-sidered as a precursor to a tentative denition of Cloud BPM (see 1.3). Aftera short linguistic prelude, business process management (BPM) as a disciplineis discussed, followed by a brief look at how BPM is put into practice usingBPM software tools. Then, cloud computing is considered in its generic as-pect. Finally, consideration is given to how Cloud BPM is viewed by analysts,software vendors, and BPM commentators and practitioners. It is through thesynthesis of this information that a denition of Cloud BPM is then formu-lated, as presented in Chapter 4.2.2 Cloud BPM from the linguistic point of viewCloud BPM is a compound noun comprising two elements: cloud, whichrefers to the notion of cloud computing, and BPM, which in this case refersto the software tools used in support of the management discipline knownas business process management. While business process management is amature, well-dened concept, cloud computing is less so. Although its originscan be traced back to computing concepts espoused in the 1960s (Hugos andHulitzky, 2010; Wardley, 2009), cloud computing is a rapidly evolving concept,as it incorporates the rapid advances of the technologies that support it, notto mention the cultural shift that signals its ever wider adoption. Therefore, ifCloud BPM is a term which is in need of a denition, it is so largely becauseof the imprecision involved in the component term cloud. Nevertheless, the term BPM also has some ambiguities attached. It isimportant to distinguish two dierent usages of the term. On the one hand, 6 17. 2. LITERATURE SURVEYthere is BPM the management discipline; on the other, there is BPM the technology, the means by which BPM is implemented in the organization(Viaene et al., 2010). It is clear that cloud (computing) denotes a type of technology, so whencloudis combined with the term BPM to yield Cloud BPM, it is under-stood that BPM in this case refers to the technology by way of which BPMis implemented, and that the technology in question is cloud based. Notwithstanding the particular case of the term Cloud BPM, wheneverthe technology of BPM is intended (and not the discipline), the term businessprocess management system (BPMS) is commonly used, and that is the usagethat is employed in what follows here. The analysts Gartner have in the pastused the term business process management technology (BPMT) to referto the software element of BPM, but now generally use the term businessprocess management suite (BPMS), which implies a comprehensive BPMsoftware package that provides a standard range of functionality (modelling,deployment, execution, etc.) (McCoy, 2011). For the purposes of this project,these two meanings of BPMS business process management system andbusiness process management suite can be considered synonymous.2.3 Business process managementBPM as a management discipline has its origins in previous management dis-ciplines such as business process reengineering (BPR), as developed in theseminal works of Hammer and Champy in the 1990s (Ko, 2009), and TotalQuality Management (TQM) (Viaene et al., 2010). Ko (2009) also cites Dav-enports seminal contribution in emphasizing the crucial role of informationtechnology in the implementation of BPR in particular.2.3.1Dening BPMIn order to understand what BPM is, it is tting to begin with an appre-ciation of what is meant by a business process. Weske (2007, p5) denes abusiness process as a set of activities that are performed in coordination in anorganizational and technical environment in order to realize a business goal.According to Weskes denition of the term, each business process is enactedby a single organization [emphasis added], but it may interact with businessprocesses performed by other organizations (loc. cit.). 7 18. 2. LITERATURE SURVEY Bearing in mind this denition of a business processes, business processmanagement can now be dened as: supporting business processes using methods, techniques and soft- ware to design, enact, control and analyze operational processes in- volving humans, organizations, applications, documents and other sources of information (van der Aalst et al., 2003).So, BPM is a management discipline which may include the use of softwaresystems to support its aims. In most organizations today, the management of business processes involvesthe use of a software platform to orchestrate a combination of both automatedand human tasks. Such business processes are referred to as executable businessprocesses (Crusson, 2006). The generic software systems which use explicitprocess representations to coordinate the enactment of business processes aretermed business process management systems (Weske, 2007, p6).2.3.2 The BPM lifecycleIn order to understand both BPM and the technologies which support it, anunderstanding of the BPM lifecycle is necessary. van der Aalst (2004) identiesa BPM lifecycle comprised of four stages (see Figure 2.1), as summarized hereby Ko et al. (2009): Process design. As-is business processes are modelled in the BPMS. System conguration. The BPMS and the underlying system infrastruc- ture is congured. Process enactment. Electronically modelled business processes are de- ployed in BPMS process engines. Diagnosis. Using analysis and monitoring tools, ow times, process bot- tlenecks, utilization, etc. can be identied and improvements suggested. The BPM lifecycle may be preceded by other steps, for example, before theprocess design phase there is the necessary step of process discovery, whichcan involve the collaboration of many dierent stakeholders in dening theprocesses to be modelled. However, once the lifecycle is initiated, the analysisphase normally will lead back into the design phase in which process improve-ments suggested in the diagnosis phase can be implemented. 8 19. 2. LITERATURE SURVEYFigure 2.1: van der Aalst et al.s BPM lifecycle (Ko et al., 2009)2.3.3 The BPM disciplineThe aim of business process management is to improve the business perfor-mance of an enterprise by changing business operations to perform more ef-fectively and eciently (Samarin, 2009, p1). A key characteristic of managedprocesses is that they are adaptive, that is, information derived from the di-agnosis of the processes is used to adjust and optimize the process in its nextiteration. This concept of continuous improvement (Palmer and Mooney,2011) is inherited from other management disciplines such as Total QualityManagement, Lean Management and Six Sigma, but extends the concept toinclude management all types of business processes, across the enterprise andbeyond.2.3.4 BPM technology: the BPM suiteA BPM system (BPMS) oers agility and exibility to enterprise softwaresolutions, in contrast to traditional enterprise software, which was designed toprovide process optimization through standardization (Jost, 2011). A BPMsystem provides a business process abstraction layer over an organizationsapplications and software services (Hill and Sinur, 2010). Process centredBPM initiatives change the entire notion of a business application becausea BPM enabled application responds to process context rather than routingprocesses around the limits of technology (Palmer and Mooney, 2011). Onceagain, BPM initiatives put process rst (Ould, 2005).9 20. 2. LITERATURE SURVEY A BPMS can support the entire lifecycle of business process development discovery, modelling, execution, monitoring, optimization from design-timeto run-time (Kemsley, 2011c) (see Figure 2.2 below). BPMSs provide a com-position environment and process modelling tools to graphically reassembleexisting functionality outside the suite (usually in the form of services madeavailable through the implementation of a service oriented architecture) tocreate a process application. A registry and repository are required to locateWhat is a BPMS?these reusable assets in the form of services (ibid.) (see 2.3.5). Performance Management - Dashboards Integration - Analytics adapters - BAM Performance Data BusinessSystems ERP Integration Framework Process Design CRM Process Modeling - Flow - Flow- Resources Process EJB - Resources/costs - DataEngine - KPIs- Business rules BusinessLegacy - Simulation analysis - FormsRules - Integration Business ITHumanUser User UserUser workflowFigure 2.2: Components of a BPMS (Silver, 2006) According to Linthicum (2009, p129), the other components of a BPMtechnology solution are: a business process engine that controls the execution of a process andmaintains the state of each of the process instances, a business process monitoring interface [performance management] forthe monitoring and optimization of processes, a business process engine interface that allows the other applications toaccess the business process engine, and10 21. 2. LITERATURE SURVEY integration technology that is required to enable the various systems and services to communicate.In some cases the integration function may be performed via existing mid-dleware (e.g. an enterprise service bus) external to the BPMS, in others, viaintegration technology bundled with BPMS solution itself. According to the analysts Gartner (Hill and Sinur, 2010), a BPMS servesto support the following key aspects of the BPM discipline: optimizing the performance of end-to-end business processes that span business functions, as well as processes that might extend beyond the enterprise to include partners, suppliers and customers making the business process visible (i.e., explicit) to business and IT constituents through business process modelling, monitoring and opti- mization keeping the business process model in sync with process execution empowering business users and analysts to manipulate a business process model to modify instances of the process enabling the rapid iteration of processes and underlying systems for con- tinuous process improvement and optimizationOne can see that a prominent feature of the BPMS is its business centredfocus; it is a comprehensive tool that is intended to support the entire BPMlifecycle from design, to deployment, to analysis and optimization. Fromthe business users point of view, perhaps the most important function thatthe BPMS oers is to provide operational transparency by making businessprocesses visible (Gilbert, 2010). According to Gartner (Hill and Sinur, 2010), the top four usage scenariosthat drives companies to invest in BPMS are: support for a continuous process improvement program implementation of an industry-specic or company-specic process solu- tion support for a business transformation initiative 11 22. 2. LITERATURE SURVEY support for a process-based, service-oriented-architecture (SOA) redesignIn order to support these various use cases, BPMSs must provide comprehen-sive functionality and the ability to integrate with the rest of the technologystack. In terms of processes, a BPMS must be capable of handling all of thefollowing (Kemsley, 2011c): straight through processes (fully automated) long running with human input dynamically changing process ows collaboration within processesBusiness processes cover a wide spectrum, from structured, repeatable pro-cesses to unstructured dynamic processes and case management, but processesusually comprise a mixture of types (Kemsley, 2011b) (see Figure 2.3). Giventhese dierent types of processes, new products are evolving and dierentiat-ing according to the type of process they focus on. Thus, we now see CaseManagement software emerging as a separate category of BPM, which focuseson the management of long running, unstructured document based processes,which comprise a series of human tasks.2.3.5BPM and service oriented architectureAs has been mentioned, the building blocks of business processes consist ofservices, well-dened blocks of functionality that are available to be orches-trated into a business process. The availability of such services is dependentupon systems which are architected in such a way that services, these discreteblocks of functionality, can be located and consumed. This is accomplishedby way of a service oriented architecture (SOA). Linthicum (2009, p5) denesservice oriented architecture as: a strategic framework of technology that allows all interested sys- tems, inside and outside of an organization, to expose and access well-dened services, and information bound to those services, that may be further abstracted to process layers and composite appli- cations for solution development.12 23. 2. LITERATURE SURVEYFigure 2.3: Spectrum of business processes (Kemsley, 2011b)The exposure of these well-dened, loosely coupled services is accomplished viainterfaces which rely on common interface denition languages (Papazoglou,2008; Weske, 2007). Business users dene the processes they need to imple-ment, and the BPM system (with or without the intervention of the technologyteam) identies the services that are required in order to supply the neededfunctionality. SOA is the means whereby these services are made available forimplementation. Ideally, the two concepts should be bridged into the sameplatform but in the meantime developers can use Web Services platforms towrap existing application adapters and expose them to the BPMS (Crusson,2006) (see Figure 2.4). In the classical service oriented architecture, a service provider publishesa service to a service registry. The service requestor then requests a servicefrom the service registry, which in turn replies with the information necessaryto allow the service requestor to bind with the nominated service provider(Weske, 2007, p59).13 24. 2. LITERATURE SURVEY Figure 2.4: BPM and SOA (adapted from Crusson, 2006)2.3.6 BPM adoption and potential obstacles to be overcomeBPM is certainly being marketed as an important trend in business software.According to Ko et al. (2009), as early as 2006, research by Gartner found BPMsystems to constitute a mature, established middleware product oering thatwas predicted to sustain 24% annual growth in the market. However, manycommentators have commented on the slow rate of BPMS adoption (Dubray,2007; Patig et al., 2010; Spurway, 2011), citing both functional and technicalfactors in explanation. Spurway (2011) in particular accuses the BPM industryof over-hyping the simplicity of BPM tools and the extent to which businessusers can easily implement BPM solutions without the need for substantial ITsupport in the overall process. Deane (2011) similarly disputes the reality ofa comprehensive business process solution that eectively bypasses IT.1 Silver(2006) characterizes BPMS as neither business user centred nor the means fora clean hand-o from business to IT, but rather, a means whereby businessand IT can collaborate on a process throughout the BPM lifecyle. This isperhaps the correct view, provided that the business process model being used 1 However, there is much anecdotal evidence that BPM-as-a-service solutions especiallydo provide an opportunity for line of business implementations of BPM initiatives, perhapsin the form of pilot projects, or simple, domain specic needs, and this aspect of cloud basedBPM may be an important factor in its favour. 14 25. 2. LITERATURE SURVEYcan be easily shared and understood by both parties.1 The current standard for BPM modelling is BPMN 2.0 (Business ProcessModelling Notation). The goal of BPMN isto provide a business process modeling notation that is readily us-able by business analysts, technical developers and business peo-ple that manage and monitor these processes. One of the goal ofBPMN is also to be able to generate execution denitions (BPEL4WS)that will be used to implement the business processes. As such,BPMN positions itself as a bridge between modeling and execu-tion and between people that run the business and implementersof systems that support the business. (Dubray, 2004)Many commentators have questioned the degree to which BPMN 2.0 is actuallyaccessible to average business users (rather than specialists, such as businessanalysts), and its ultimate suitability for the modelling of executable processesof any degree of complexity (ebizQ, 2011). Another obstacle in the path of BPM adoption using BPMN is the round-tripping problem, as described by Silver (2007):A process model created in BPMN or comparable owchartingnotation could not be easily kept in sync with the executable BPELdesign throughout the implementation lifecycle. Essentially, youcouldnt update the process model from the BPEL.. . . So the modelwas not a continuous business view of the implementation. In fact,it was still what it had always been initial business requirements.Some vendors tried to bypass this problem by focusing on human-centric pro-cesses, leading to a new style of BPMSin which executable design is layered directly on top of the processmodel, in the form of implementation properties of BPMN activ-ities. The new style does not create a hando between dierenttools (with dierent ow models, data models, and programmingmodels), but leverages a single tool shared by business and IT,with business focused on the activity ow and IT focused on mak-ing those activities executable. (Silver, 2007) 1 This round-tripping problem is discussed below. 15 26. 2. LITERATURE REVIEWHowever, this solution was a partial one at best. Dubray (2007) suggestedthat (as of 2007) no vendor could claim that a general purpose engine thatbusiness analysts can use (even with minimal intervention from IT) to createa solution from process models had been delivered. Agreeing with Silver,he claimed that the limited success that vendors had achieved was the resultof the focus on human-centric processes, which for the most part t wellthe centralized view of a business process engine developed by these vendors,especially when limited customization of and integration with existing systemsis needed (ibid.). In summary, problems with the complexity of the BPMN modelling no-tation, coupled with the diculties relating to the translation of models intoexecutable code (using BPEL) meant that vendors were faced with an under-standable resistance to widespread adoption of BPMSs. Indeed, based on the responses of over 130 Forbes 2000 Global companies,a recent study by Patig et al. (2010) showed BPM adoption to be at a lowerlevel than what might be expected. The authors found that BPM maturityin most companies was at a low to intermediate level, and cited the lack ofBPM in a SaaS format as being one possible factor contributing to the lackof adoption, with the over-complexity of bundled BPM modelling tools beingsuggested as another. Although the adoption of BPM products continues to advance, the extentto which business processes are utilizing cloud based services is still quite low;a Gartner survey conducted in 2010 found that only 40% of companies withBPM systems had even a small proportion (10%) of their processes utilizingservices based in private or public clouds (Gartner, 2011).2.4 Cloud computingIt is dicult to agree on a comprehensive denition of cloud computing, asit is a technology which supports a wide variety of use cases. As a general-ization, Wardley (2009) characterizes cloud computing as a disruptive shiftof the computer stack to online services, allowing on-demand access to soft-ware applications, development and deployment environments, and computinginfrastructure, on a pay-per-usage basis. 16 27. 2. LITERATURE REVIEW2.4.1Dening cloud computingA more comprehensive denition has been proposed by the National Institutefor Standards and Technology (NIST, 2011): Cloud computing allows computer users to conveniently rent ac- cess to fully featured applications, to software development and deployment environments, and to computing infrastructure assets such as network-accessible data storage and processing.Some observations on this denition are in order. Firstly, as has become con-ventional, cloud computing is here dened as comprising three service models(see Figure 2.5):1. Software-as-Service (Saas). An application that is hosted and delivered to the customer by a software provider.2. Platform-as-a-Service (PaaS). A development environment where a cus- tomer can create and develop applications on a providers computing environment.3. Infrastructure-as-a-Service (IaaS). An o-premise data centre environ- ment. (ibid.)Figure 2.5: Cloud computing (Source:http://contactdubai.com/web hosting/advantages-of-cloud-computing) 17 28. 2. LITERATURE REVIEWSecondly, cloud computing services are rented, that is, the service is providedand charged for on the basis of usage, either based on a subscription model,or on actual usage, such as compute cycles consumed or data stored. Thirdly,the services can be conveniently accessed, which eectively means by wayof any standard web browser. NIST (2011) go on to caution that a comprehensive denition of cloudcomputing is not possible, due to the fact that cloud computing is not asingle kind of system, but instead spans a spectrum of underlying technologies,conguration possibilities, service models, and deployment models. However,the following ve characteristics are identied:1. On-demand self-service. The service can be accessed by the user, as and when required.2. Broad network access. The service can be accessed from a variety of devices, using standard network protocols.3. Resource pooling. The service uses a multi-tenant model, using a combi- nation of physical and virtual machines, assigning resources dynamically according to user demand.4. Rapid elasticity. From the customers point of view the service can be scaled up or down on demand, as needed.5. Measured service. Resource usage is monitored for the purposes of billing, as well as for service quality purposes. (ibid.) Finally, four deployment models are identied:1. Private cloud. The cloud infrastructure is for the sole use of the organi- zation, although it may be managed by a third party and may be located o-premise.2. Community cloud. As above but with the infrastructure being shared by a number of organizations with shared concerns.3. Public cloud. The infrastructure is owned and operated by a provider and made available to the public.4. Hybrid cloud. A combination of two or more of the above types of clouds which remain distinct, yet are bound together by standardized 18 29. 2. LITERATURE REVIEWor proprietary technology that enables data and application portability(e.g. cloud bursting for load balancing between clouds). (ibid.) In what may be seen as an indication of the rapid evolution and maturationof the cloud computing paradigm, Linthicum (2009) proposes the to catego-rization of cloud computing with increased granularity, thereby identifying 11major categories: 1. Storage-as-a-service 2. Database-as-a-service 3. Information-as-a-service 4. Process-as-a-service 5. Application-as-a-service 6. Platform-as-a-service 7. Integration-as-a-service 8. Security-as-a-service 9. Management-as-a-service 10. Testing-as-a-service 11. Infrastructure-as-a-serviceIt should be noted that process-as-service here refers to ready to use processes,that is, a set of orchestrated services that can be consumed as a component tobe further orchestrated into a larger process; it does not refer to a platform forthe the construction and management of business processes (a BPMS). MostBPM cloud oerings are classied as Software-as-a-service or Platform-as-a-service.2.4.2Characteristics of cloud computingArchitecture. Insofar as the cloud is preeminently designed to provide ser-vices, the cloud shares with SOA a common foundation (see 2.3.5). Accordingto Rosen (2011), the same service design principles that make a good SOAservice need to be applied to a cloud service: well dened interfaces, loosecoupling, proper decomposition, common semantics, etc.19 30. 2. LITERATURE REVIEWUse cases. Armbrust et al. (2010) identify three main use cases for cloudcomputing: (1) when demand for a service varies with time, (2) when demandis unknown in advance, and (3) for batch data analytics requiring short burstsof high resource usage.Advantages. The primary economic advantage of cloud computing is a re-sult of the elasticity of the service: because the service is made available ondemand, there is no need to over-provision resources to meet peak demand;similarly, there is no risk of under-provisioning and the resulting loss of rev-enues due to the inability to meet demand (ibid.). Elasticity refers to theability of the service to expand or contract resources in the very short term almost instantaneously according to load. Consumers pay for the resourcesthey are using, unlike on-premise resources which need to always maintain thecapacity to cope with peak demand levels. Another advantage of the cloud model is scalability. With reference tocloud services, scalability is often used synonymously with elasticity, however,scalability may also refer to the ability of a cloud based service to facilitatean expansion of business operations, not in the moment, but rather, overtime. Whereas elasticity refers to the responsiveness of resource allocationto workload, and is characteristic of shared pools of resources, scalability isa feature of the underlying infrastructure and software platforms (Gartner,2009).Concerns. Cloud computing necessarily involves trusting ones data to athird party, so in addition to there being the normal concerns about datasecurity that are applicable in any context, there are also concerns about towhat degree the provider is capable of guaranteeing security, especially whentheir arrangements may involve other third parties, for example, the utilizationof storage provision from other cloud providers. Cloud customers also need to be aware of where their data might be stored,as that has implications for data privacy, which is dependent upon the lawsof the country under whose jurisdiction the stored data falls. According toKemsley (2011, pers. comm., 12 June), many companies are reluctant toput their processes in the cloud because of the potential for not only securitybreaches, but also government intervention in the data. Finally, cloud computing admits of all of the security concerns inherent20 31. 2. LITERATURE REVIEWto large-scale systems, especially those that depend on virtualization tech-nologies that are still not well understood (Chorafas, 2010, p206). Of course,much of the cloud computing model is heavily dependent upon virtualizationtechnology.2.5 Cloud BPMIn the preceding sections the two constituents of Cloud BPM were consideredis some detail. In the following section the focus returns to Cloud BPM proper,beginning with a brief glimpse into its the origins.2.5.1 HistoryAccording to Weske (2007, p25), business process management mainly dealswith information systems in the context of enterprise systems architectures.Today, the nature of this architecture is changing as the needs of business,and the nature of the business processes that support business, are changing.The latest development in this evolution of enterprise systems architectureis the growing adoption of cloud computing technologies, as described above(Section 2.4). Cloud BPM, the marriage of BPM and cloud computing tech-nologies, is the logical outcome of the need to manage business processes inthis new context of cloud computing. The more processes move o-premise,the more compelling the argument for cloud based BPM becomes (CordysB.V., 2011a). In earlier times, although in a free market economy services should bemost economically provided externally, for corporations operating at scale,most services could be provided more eciently in-house, saving the additionalexpenses that would be incurred in locating, contracting, coordinating andpaying for such services from external sources (Hugos and Hulitzky, 2010, p1).Today, however, because of the technological advances that aord increasedconnectivity through web services and the like, it is increasingly the case thatservices can be procured more economically outside the enterprise. Thus,the modern enterprise itself has metamorphosed into a new incarnation, thatof the virtual enterprise (Hugos and Hulitzky, 2010). These developmentsinitially led to the outsourcing of complete business processes, but today, asthe services on oer become increasingly granular and accessible, more controlcan be gained by orchestrating these services to compose business processes in21 32. 2. LITERATURE REVIEWthe form of custom composite applications that are exible, agile and adaptive(see Dubray, 2008). The rst glimmerings of Cloud BPM emerged around 2006 or so, whenvendors began to oer modelling tools that would run in a browser and beaccessible on the internet, announcing these as BPM platforms oered asa service. However, these were not considered as tools t for any seriousBPM purpose (Ghalimi, 2007). Writing in 2006, Khan stated that a TrueBPM-as-SaaS oering should be capable of managing and executing com-plex, personalized, fully-integrated, mission-critical processes and have theability to adapt the processes on the y to meet changing business condi-tions. In stark contrast, he noted that the BPM-as-SaaS oered at that timewas restricted to providing partial functionality, such as modules providingmodelling or documentation, templates oering simple pre-dened processes,or simple hosting. Both authors were implying that SaaS BPM should includethe ability to execute processes. He also identied a lack of exibility as beingone of the limiting features of SaaS in general and puts this forward as beinga challenge for vendors wishing to provide a BPM-as-SaaS. By 2009, bloggers were beginning to ask what BPM in the cloud was,and what it might be good for, writing articles with titles such as BPM andcloud computing (Silver, 2009) and BPM in the Cloud: one plus one ismore than two (Byron, 2009a). Some of these articles are discussed below,in Section 2.5.4. By 2010, ten of the top 15 BPMS vendors (based on worldwide total BPMSsoftware revenue in 2009) were oering cloud-enabled BPMS platforms (Gart-ner, 2010), with a much larger number of smaller vendors adding to the mixof options available. Some of these oerings are discussed in the followingsection.2.5.2Vendor oeringsIn this section, overviews of a small sample of Cloud BPM products are given.The products mentioned are from the vendors Appian, BonitaSoft, Cordys,and Intalio. 22 33. 2. LITERATURE REVIEW2.5.2.1AppianAppian is one of the top vendors of BPM oering a cloud enabled BPMS.A web search on Cloud BPM consistently has Appian Cloud BPM as therst result, and this has been the case for the duration of this project (June September 2011). Their choice of the generic-sounding Cloud BPM as thename of their oering appears to have been a good choice in search engineoptimization terms. Appians Cloud BPM webpage mentions decreased costsas the principal benet of this deployment method, and goes on to tout Ap-pians security and reliability features. Clearly, Appian think that customersare looking for cost savings (as well as rapid deployment timeframes), and aremost concerned about the security and reliability of cloud deployment.Appian oers a cloud deployment that aords the same functionality astraditional on-premise BPM software deployments (Appian Corp., n.d.), in-cluding easy SOA integration using packaged connectors for common systeminterfaces and native support for SOA frameworks (Appian Corp., 2011a).Appians BPM products include SaaS and PaaS oerings, with PaaS beingthe more popular option with customers, according to Samir Gulati, vice pres-ident of marketing for Appian (All, 2011). In many cases, customers choosethe cloud deployment as a way of expediting the BPM implementation, andswitch over to an on-premise solution once the pilot project is working well(ibid.).2.5.2.2BonitaSoftBonita Open Solution is a BPM system using open source technology to pro-vide a fully featured BPM product including a BPMN modelling tool, a BPMand workow process engine, and an advanced, clean user interface (see Fig-ure 2.6) (BonitaSoft, 2011). The Bonita Studio modeller allows users to choosebetween a simple or advanced palette, thus determining which subset of BPMN2.0 features are made available to the user. The modelling component includesover 100 built-in connectors to build processes that include services derivedfrom a number of commonly used commercial and open-source databases,ERPs, CRMs, etc., and also includes process simulation. The monitoringcomponent features custom dashboards and reports using custom-dened keyperformance indicators (KPIs), as well as real-time activity monitoring.23 34. 2. LITERATURE REVIEWFigure 2.6: Bonita Studio (taken from BonitaSoft (2011))2.5.2.3 CordysCloud based BPM represents only one aspect of Cordys grander vision ofcreating a comprehensive, cloud based enterprise software platform, utilizingstate-of-the-art technologies to enable enterprise systems that deliver the fea-tures and performance that the enterprises of today require. The cloud basedplatform comprises three main functions: integration, BPM, and compositeapplication development (see Figure 2.7). The Cordys Business Process Management Suite forms one of the maincomponents of a comprehensive BPM platform called Cordys Business Op-erations Platform (BOP-4) (see Figure 2.8), which allows for the design, ex-ecution, monitoring and continuous optimization of business processes, andincludes components such as Business Activity Monitoring (BAM), MasterData Management (MDM), Composite Application Framework (CAF), andSOA Grid (ESB). Cordys also oers a more lightweight platform, the CordysProcess Factory, which allows SMB or departmental users to build and runprocess-centric mashup applications on the Web. The Cordys BPMS appears to have solved the round-tripping problemmentioned above; Cordys claims that the platform allows business and IT to24 35. 2. LITERATURE REVIEWFigure 2.7: Cordys cloud platform(Source:http://www.cordys.com/cordyscms com/improving business operations.php)work on the same process model, which always stays in sync (Cordys B.V.,2011a). It is a completely browser based product which features highly respon-sive AJAX based applications and oers enterprise grade scalability, reliability,security and standards support. Some of the features of the Cordys BOP-4 architecture are described belowand indicate to what extent Cordys BPM platform is optimized for clouddeployment. AJAX based applications on browser Modelexecution synchronization Browser based collaborative workspace Stateful objects and stateless connections for near-linear scalability Integration-ready SOA for both internal and external interfaces High availability with Cordys clustering technology (State Synch-up) Reliable transport support (JMS, MSMQ) Standardized on WS-Basic prole compliance, WS-Security support, etc. 25 36. 2. LITERATURE REVIEWFigure 2.8: Cordys Business Process Operations Platform (Source:http://www.cordys.com/cordyscms com/platform overview.php)26 37. 2. LITERATURE REVIEW Pluggable and loosely coupled architecture internal component com- munication also uses Web services The Cordys platform has the ambition and vision to supersede the data-centric ERP systems of decades past with an Enterprise Cloud Orchestrationplatform that is process-driven, to provide agile and responsive solutions torapidly changing business requirements (Cordys B.V., 2011b).2.5.2.4 IntalioIntalios cloud based BPM suite is called Intalio|BPM and is legacy-freesoftware, a purpose-built cloud application. It oers 100% Web-based user interfaces, native multi-tenancy, a small memory footprint optimized for virtualization, and support for the most popular deployment options, including VMware vCloud, Mi- crosoft Azure, and Amazon EC2, both on premises and on demand. (Intalio, Inc., 2011)The latter deployment options would enable single-tenancy deployments, whichsome would consider to oer greater security of data. Intalio|BPM is a full feature BPM suite. Its features are organized acrossa twelve-step cycle for business processes, from process discovery to processcontrol (Intalio, Inc., 2011), including modelling, simulation, execution, mon-itoring and analysis. Intalio|BPM is architected such that its next-generationprocess engine is capable of executing BPMN 2.0 processes natively, withouthaving to resort to any code translation (Intalio, Inc., 2011) into BPEL orotherwise (see Figure 2.9). Intalio|BPM also supports complex workow pro-cesses and Adaptive Case Management scenarios, includes a fully extensibleHuman Task Manager service compliant with the WS-HumanTask industrystandard. This service implements the end-to-end life cycle for human tasks,and can be easily modied to support custom steps and transitions, whiletaking full advantage of a powerful built-in Business Rules Engine (BRE).2.5.2.5 Other vendorsTibco consider that BPM in the Cloud promises increased IT eciency, re-duced capital expenditure, and lower barrier to entry, while providing scala-27 38. 2. LITERATURE REVIEWFigure 2.9: Intalio|BPM Architecture28 39. 2. LITERATURE REVIEWbility with innite computing power (Tibco Software Inc., 2011). Pegasys-tems (Pegasystems Inc., 2011a) oer in their cloud BPMS product all of thefunctionality of their on-premise BPMS, as well as promising unparalleledsecurity and reliability and out-of-the-box integration to existing data cen-ters. They also mention an advantage of the cloud deployment, claimingthat multi-enterprise processes are ideal for cloud deployment, automatingthe interactions between multiple parties and ensuring that SLAs are fullled(ibid.). But Pegasystems Founder and CEO, Alan Treer, acknowledges thereluctance of some users to trust their highly strategic processes and data tothe cloud: Business users have become increasingly intrigued by the SaaS model, but have told us they are not going to trust their mis- sion critical processes, policies and data to an externally hosted environment. (Pegasystems Inc., 2011b)2.5.3 Analyst point of viewIn July 2010, the analysts Gartner characterized cloud enabled BPM asan emerging technology with potentially high benets but with low marketpenetration to date (Gartner, 2010). Cloud enabled BPM is dened as soft-ware that use BPM technologies to construct and optimize process-centricsolutions in a software-as-a-service or cloud service delivery model, technolo-gies including high-level process modelling tools, business process analysissoftware, workow, automated business process discovery tools, BPM suites,business activity monitoring, and business rules management systems (ibid.). Cloud enabled BPM solutions may be provided as a platform-as-a-serviceor embedded in software-as-a-service solutions. Two common use cases forCloud BPM mentioned are as platforms for collaborative modelling of busi-ness processes, and the adoptions of BPMSs for BPM pilot projects. Perceivedbenets mentioned are cost savings and scalability, especially for midsize com-panies who may not otherwise be able to acquire this technology. Cloud en-abled BPM is also seen as enabling increased collaboration in BPM projects: Gartner believes that extreme collaboration is critical to impacting change and improving performance. Cloud computing accelerates collaboration and allows BPM and SOA initiatives to have an even greater impact. (Software AG, 2011)29 40. 2. LITERATURE REVIEW The analysts Forrester mention the value of cloud computing to accelerat-ing delivery and minimize risk (Kemsley, 2011a). BPM-as-a-service will lower barriers to getting started with BPM suites. Startup costs for implementing BPM suites can put these tools out of reach for some process owners that dont already have budget and executive support for launching their process initiative. To prove initial value of BPM suites, smart process professionals now turn to BPM suites hosted in the cloud often referred to as BPM-as-a-service.2.5.4 Practitioners, bloggers, commentatorsVendors and analysts both have their views regarding cloud enabled BPM, andthese two groups can display a degree of symmetry in their outlook. Whatis really important to the future of cloud based BPM, however, is how thistechnology is viewed by practitioners of BPM. The source of reference for theseopinions is a number of weblogs and the discussion that they spawn. In thesediscussions the true state of Cloud BPM can be discerned what Cloud BPMis, what is promises, what it lacks in short, the issues that the BPM customerfaces when contemplating the purchase and deployment of cloud enabled BPMsystems. Much of this discussion occurred around 2009, when the Cloud BPMoption was just beginning to become available. Wainewright (2009) was one of the BPM consultants early on hinting atthe possibility that cloud BPM could oer a new future for BPM. Commentingon the ebizQ article How does using a BPM solution in the cloud dier fromusing an on- premise BPM application? Which is better?, he wrote: However one might also ask whether, looking further ahead, a cloud environment would ultimately change the nature of BPM because of factors such as easier modication, more standardized integra- tion and APIs, and the ability to do more process integration at the user interface layer by taking advantage of standardization on web client technologies such as the browser, AJAX, Flex and so on. Also writing in 2009b, Byron canvassed the views of BPM practitionerswith his ebizQ article, Calling for input on BPM in cloud computing: lets30 41. 2. LITERATURE REVIEWclear away the fog. Having done his sums, the results were written up in anarticle entitled BPM in the Cloud: one plus one is more than two in whichhe states:One interesting thing about a BPM in the Cloud architecturalanalysis is that the basic design of the BPM-enabling software(or any other type of software in the cloud) could make a dier-ence. Presumably software is more functional if it is designed orre-designed to run in the cloud as opposed to simply taking advan-tage of the clouds characteristics.Byron goes on to substantiate this claim by citing the view of one vendor,Software AG, that the cloud lets BPM analysts and developers more easilycollaborate on process discovery (gathering artefacts, nd out who does specicwork, identify who the process expert is, etc.) (ibid.). In this view thebeginnings of the current focus on social BPM is evident. Khoshaan (2011) notes the robust relationship between Cloud BPM so-cial networking. Business processes provide the context of collaboration, andsocial networking supports and augments the various phases of the BPM con-tinuous improvement lifecycle. Barlow (2009) points out that cloud BPM platforms provide all the advan-tages that traditional SaaS oerings such as CRM and workforce managementsystems provide, without the expected drawback of reduced exibility. SinceCloud BPM is a platform-as-a-service, the system is can evolve functional-ity through the creation of process-oriented business applications, rather thanmerely utilizing the limited, built-in functionality that SaaS software provides. Sandy Kemsley (2011, pers. comm., 12 June), a prominent BPM consul-tant, cited security/privacy concerns as the biggest issue with cloud BPMcurrently. Vendors acknowledge these concerns too, Appian for example.Cloud computing promises lowered IT costs and faster time-to-value thantraditional on-premise deployments, but the cloud model is still new terri-tory and many questions particularly around issues of data security persist(Appian Corp., 2011b). 31 42. 2. LITERATURE REVIEW2.6SummaryIn this chapter, the literature surrounding the concept Cloud BPM has beenreviewed.1 This began with a discussion of each of the two elements of theCloudBPM marriage cloud computing and business process management.Thereafter, Cloud BPM, as it has developed since around 2006 until thepresent, and as evidenced by the views of vendors, analysts and BPM practi-tioners writing on the internet, was discussed. Certain themes have emerged,and these will inform the tentative denition of Cloud BPM that is proposed,and then tested, in the chapters following. 1 For the sake of completeness, one other manifestation of Cloud BPM should be men-tioned. Linthicum (2009, p127 .) discusses the relocation of information, service andprocesses [emphasis added] to the cloud, rather than the relocation of a BPMS to thecloud, and is therefore invoking the concept of BPM-as-a-service, mentioned above in Sec-tion 2.4.1.32 43. 3. Methods3.1IntroductionThe purpose of this research project, as mentioned previously in Chapter 1,was to investigate, characterize and dene Cloud BPM. The project followeda sequential process, consisting of four main tasks: (1) review the eld, (2)build a theory, (3) test the theory, and (4) reect and integrate. The subtasksfor each of one these tasks are shown in Figure 3.1 below. The researchdata for the project was generated via the completion of two main tasks: theliterature survey and the online survey questionnaire, which are described inthe following two sections. Figure 3.1: Cloud BPM project process33 44. 3. METHODS3.2Literature surveyThe rst phase of the project (the Review phase) was to conduct a compre-hensive literature survey of the project domain. The results of the literaturesurvey have been presented in the the previous chapter (Chapter 2). Thepurpose of the literature survey was twofold: (1) to gain familiarity with theproject domain, its issues and dening features, and (2) to gain an under-standing of the key aspects of the domain, namely, the theory and practice ofbusiness process management and cloud computing, all as a precursor to theprojects aim of dening Cloud BPM. This understanding would providethe foundation upon which a proposed denition of Cloud BPM would bebuilt.3.2.1Literature searchThe rst aspect of the literature search involved an intensication of the meth-ods that the author had already employed in developing an interest in thedomain of the project in the rst place; thus, the information provided byindustry analysts and bloggers on the Internet was reviewed, and referencesto further articles were investigated. Twitter1 provided an important sourceof information from both vendors and commentators. The author followedon Twitter a number of prominent BPM and technology commentators andanalysts, as well as vendors, in order to be alerted of current discussions, up-coming webinars or new products or features. A partial list of these is providedin Appendix E. The websites of a number of providers of BPM technologieswere consulted, and many of them provided links to relevant whitepapers, aswell as slide presentations and previously recorded webinars. Background research about BPM and cloud computing in general was con-ducted by way of library searches, using the City University library website2 ,and the British Library website3 , where a number of books and journals werelocated and consulted. 1 http://twitter/com/ 2 http://www.city.ac.uk/library 3 http://www.bl.uk34 45. 3. METHODS3.2.2Literature reviewOnce the main sources for the literature survey had been identied and gath-ered, the literature review was begun in earnest. The background topics wereresearched and written up, and the central topic of Cloud BPM was investi-gated and presented. The ndings of the literature review formed the basisfor the proposed denition and characterization of Cloud BPM (the Buildphase), which was then transformed into a hypothesis (as presented in Chap-ter 4). The hypothesis was then compared with the results of the online surveyquestionnaire (the Reect phase). The online survey is described in detail inthe next section.3.3 Online surveyAn online survey questionnaire was the method chosen to test the hypothe-sis (the Test phase) of the denition and characterization of Cloud BPM, assuggested by the results of the literature review just mentioned. Survey ques-tions were devised to canvass the opinions of BPM practitioners regarding thetopic domain. The target of the survey consisted of the members of BPM re-lated groups on LinkedIn1 as listed and described in Appendix B. The surveyreceived 38 responses. Because the survey elicited such a good response, and achieved a samplesize which could be considered signicant for the purposes of a qualitativesurvey, it was decided that the project would focus on an interpretation ofthe data from the survey and dispense with the original idea of canvassingthe opinions of a small number of BPM experts, with the online survey beinga backup plan to gather data if an insucient number of BPM experts wereavailable to cooperate in project. The survey provided a broad view of practi-tioners perspective on Cloud BPM, which was what was wanted. The idea ofconducting interviews with a few specic individuals would not have furtheredthis goal. The option of follow up was pursued, however, in two cases wheresurvey responses required further clarication. 1 http://www.linkedin.com35 46. 3. METHODS3.3.1Survey designThe survey was entitled Cloud BPM - a survey and was administered on-line. It consisted of 13 closed questions nine statements to be answeredusing the Likert scale, two tick box questions, and two multiple choice ques-tions and three open questions. For the sake of clarity, and ease of response,questions were grouped according to question type. Additionally, a section atthe end of the survey had questions to gather a minimum of personal infor-mation about the respondents. A survey invitation was sent to a total of 16dierent LinkedIn BPM Groups, whose members were invited to complete thequestionnaire. Respondents were advised that a summary of the results of thesurvey would be made available to them at a later date, should they wish toreceive this information.3.3.2General considerationsThe intention of the survey was to gather data about the perceptions of BPMpractitioners about Cloud BPM what it is and what advantages or dis-advantages it might have. The survey was intended as (in the words of onerespondent) a lightweight overview of current practitioners views of the topicdomain.Scope. The rst consideration was to produce a survey which covered thetopics that the literature review had suggested as salient.Delivery. The next consideration was to produce a questionnaire that wouldbe fairly easy to complete and would encourage a high level of participationfrom the target group.The online approach was deemed most appropriate asit would reach a large number of potential respondents. At the same time,targeting BPM related groups on LinkedIn would mean that respondents couldbe assumed to be knowledgeable about, and experienced with, BPM tools andmethods.1 Google Docs was selected as the vehicle for the online survey dueto the authors previous familiarity with it, as well as its simplicity of use andclean presentation. A Google spreadsheet is easily transformed into an onlineform suitable for a questionnaire. 1 Indeed, the level of insight revealed in many of the comments collected indicate thevalidity of this assumption. 36 47. 3. METHODSStyle.In order to make the results capable of easy analysis, a large numberof closed questions were used, with Likert scale type questions used where pos-sible and forming the majority of the questions to encourage a high completionrate. The closed questions were actually in the form of statements to whichrespondents would respond on a Likert-type scale, ranging from Strongly dis-agree (1) to Strongly disagree (5), with a (3) assumed to indicate a neutralor undecided position a valid response in the case of these questions and sonot to be excluded by choosing an even number of terms for the scale.The assertions were mixed, some being positive statements, others beingnegative statements, in order convey a degree of objectivity.Three open ques-tions were also set, mainly to give respondents a chance to personalize theirresponses. This was for two reasons: rstly, to draw out new insights that theother questions may not have adequately addressed, and secondly, to give therespondents the satisfaction of contributing a more personal view.Respondent information. It was considered appropriate to obtain dataon three aspects of the respondents: (1) their relationship or role with respectto BPM, (2) the size of the company they represented (if applicable), and (3)the industrial sector that their company falls under.1 The motivation behindthese questions was as follows.It was considered desirable to determine what role the respondents hadto BPM, and where they might fall within the businessIT spectrum, that is,whether they were business users (management, end users), software devel-opers (either working for customers or vendors), or working on the interfacebetween business and IT (e.g. business analysts), or any other role (e.g. aca-demic). It has become a convention to talk about the gap between businessand IT and it was thought important to assess where respondents stood inrelation to this dividing line.Especially in the case of respondents being end users, it was of interestwhat sector the company operated in, as well as the company size, in orderto gauge what types of companies were interested in Cloud BPM, and whattypes of processes Cloud BPM would be enacting.2 With regard to industrial 1It was not considered that respondents age or gender was of any signicance, nor theircountry of residence or activity, nor their number of years of involvement in the problemdomain, and so these types of information were not sought, and this served to streamlinethe process.2Business size rankings were based on the European Union denition (see 37 48. 3. METHODSsector, in addition to the using the traditional categories of primary, secondaryand tertiary sectors (corresponding to oil and gas, manufacturing, and servicesrespectively), a fourth category IT services was added, to allow a nergrained picture to emerge. All of the above questions were also intended to determine the degree ofvariety in the sample, in order to gauge whether the results could be consideredrepresentative of the target population.3.3.3Survey targetMembers of LinkedIn BPM groups were chosen as the target of the surveyfor the following reasons. LinkedIn is a social networking site that is focusedon the needs of professional business users. LinkedIn provides users with aprofessional online presence, and allows members to connect with each other(becoming connections) and follow their activities and discussions. Membersmay also join groups. LinkedIn Groups allow LinkedIn members to follow andparticipate in discussions around topics of interest specic to the group. Suchgroups can usually be joined by membership only and for this reason anyonereading and responding to posts in the group can be assumed to have a genuineinterest and some level of expertise in the subject domain of the group.3.3.4Motivation of questionsThis section describes the the rationale and motivations behind each of thesurvey questions, which were grouped into four sections.SECTION 1. The rst three survey questions, grouped under the headingDening Cloud BPM, were intended to derive a general sense of how respon-dents perceived the term Cloud BPM as it is used currently. The literaturereview had identied three main usages of the term and in the rst questionthese were suggested as possible responses. A fourth option of Other wasgiven, with a blank eld provided for the respondent to supply their ownpreferred denition. Q1.1 What do you think is usually intended by the term CloudBPM?http://en.wikipedia.org/wiki/Small and medium businesses).38 49. 3. METHODS the design and deployment of processes that utilize external web services any BPM tool (e.g. modelling) which can be accessed over the internet a BPM system that is deployed o-premise and consumed as a serviceThe third option is the most specic and relates to a comprehensive BPMsolution or suite, functionally equivalent to existing on-premise BPM suites.The second denition is more general and refers to any type of BPM tool (typ-ically, a modelling tool) that is accessible online through a browser. The rstdenition refers to a completely dierent notion, that of constructing busi-ness processes which incorporate externally sourced cloud services or servicescombined as processes, that is, BPM using cloud services rather than a BPMsoftware system that runs in the cloud.Q1.2 According to your denition of Cloud BPM, a cloud basedBPM solution must include which of the following components? process modelling process discovery process execution process monitoring process analysis process simulation Other:The next question was intended to follow on from the assumption thatCloud BPM did in fact refer to a comprehensive BPM solution consumed as aservice, and to determine what components such a solution would comprise.1In other words, what sort of functionality would the respondents require froma Cloud BPM solution? The three core components of BPM were included process modelling, process execution, process analysis as well as optionalcomponents such as process monitoring and process simulation. Other options, 1 It should be noted that of the three denitions proposed in the previous question, theauthor judged the rst to be the most prevalent in the literature, and the results of thesurvey conducted substantiated this view.39 50. 3. METHODSsuch as a process repository or process marketplace were not included for fearof including too many tick boxes, but an Other box was included to caterfor respondents who felt that other options were essential. Q1.3 Cloud BPM is simply a delivery model for BPM tools - nomore, no less. (Its not about WHAT you get, but HOW you getit.) [Likert scale] The nal question in this section was intended to test the hypothesis thatCloud BPM is no dierent from on-premise BPM, only the method of deliveryis dierent. The contrary of this would be that the nature of the cloud platformfor delivery of BPM solutions either (a) enhances or (b) detracts from the endproduct, functionally of technically. Most of the respondents agreed withthe proposition that Cloud BPM is simply a model of delivery, having noimplications on the nature of the product in itself.SECTION 2. The next section of the survey was headed CharacterizingCloud BPM and was intended to tease out some of the issues that surroundcloud based BPM solutions. Q2.1 Cloud BPM is a solution which is attractive mainly to theSMB market. [Likert scale] The rst question of this section related to whether Cloud BPM was pre-dominantly a solution that appealed to small and medium-sized businessesrather than large enterprises. The hypothesis here is that many of the bene-ts of cloud based BPM are related to the minimization of capital expenditureand initial outlay required, lowering the barrier of entry to BPM solutions. Q2.2 Cloud BPM is not suitable for the design and deploymentof complex business processes. [Likert scale] The next question sought to gauge the respondents perception of the ca-pabilities of cloud based BPM solutions by proposing that Cloud BPM is notsuitable for the deployment of complex business process. The implication isthat Cloud BPM solutions are more geared towards the creation of mashups,1or the design and implementation of comparatively lightweight processes that 1 Web applications that combine data and/or functionality from more than one source(http://en.wikipedia.org/wiki/Mashup) 40 51. 3. METHODScan make use of existing templates and built-in connectors to the requiredservices. Q2.3 New BPM initiatives pursued using a cloud BPM platformwill attract a lower level of business risk than the same projectpursued using traditional, on-premise methods. [Likert scale] This question was meant to gauge the respondents perception of whethercloud based BPM is better suited for developing new BPM initiatives eas-ily, without requiring the mobilization of a large amount of IT departmentresources in order to implement pilot projects, in other words, attracting alower level of business risk for the project. Q2.4 Cloud BPM entails serious - and in some cases, prohibitive- security risks.[Likert scale] Question 2.4 sought to gauge respondents perception of the level of se-curity risk associated with a cloud based BPM system. The question wasworded to nd out if respondents felt that security risks were considered tobe of such a degree that they might seriously impact any decision to be madeabout deploying BPM in the cloud. Q2.5 One of the main advantages of Cloud BPM is its synergywith social BPM technologies.[Likert scale] Question 2.5 sought to gauge respondents perception of the link betweencloud based BPM and social technologies that enable users to more easily col-laborate in the design processes, as well as monitor processes that are running.The hypothesis is that a cloud based BPM system is better suited architec-turally for the provision of such functionality. Q2.6 Due to its high strategic value to the organization, BPM isnot a suitable candidate for a cloud implementation. [Likert scale] Question 2.6 aimed to test the hypothesis that since the business processesthat a company runs are of high strategic importance, due to the security con-cerns associated with hosting the process information in a cloud environment,a cloud environment is not suitable for the implementation of a BPM system. 41 52. 3. METHODS Q2.7 The full benets of a cloud based BPM system will onlybe realized when the application is purpose-built for deployment inthe cloud.[Likert scale] Question 2.7 sought to gauge the respondents perception of the utility ofCloud BPM towards introducing BPM initiatives easily and quickly, perhapson an experimental or pilot basis. The hypothesis here is that cloud basedBPM solutions can be introduced and trialled at a very low cost, without theneed to purchase new hardware or software, and without relying on the ITdepartment to mobilize for this change. In other words, Cloud BPM ban puta BPM solution into the hands of the business users and allow them to pursuepilot projects for a quick win, in order to demonstrate the ecacy of BPMsolutions in general. Q2.8 The full benets of a cloud based BPM system will onlybe realized when an organizations IT stack is predominantly cloudbased. [Likert scale] Question 2.8 was intended to gauge respondents perception of whethercloud based BPM was more suited to the management of business processeswhen the rest of the IT stack was cloud based. The hypothesis here is thatcloud based BPM makes the most sense when the systems that it is interactingwith are architected specically to operate in a cloud environment. Q2.9 What are the main reasons for an organization to choosea cloud based BPM solution over an on-premise solution? (Pleasetick a maximum of FIVE reasons only.) quicker time to market lower start up costs reduced capital expenditure higher return on investment increased business agility elasticity of service reduced total cost of ownership 42 53. 3. METHODS scalability of service better process collaboration OtherThe last question in this section suggested some of the possible advantagesassociated with cloud based BPM, and requests that the respondent chooseup to ve main reasons. This question sought to identify the features of cloudbased BPM that respondents considered to be the most important.SECTION 3. This section consisted of two open questions and was in-tended to give respondents a chance to express their own views about CloudBPMs advantages and disadvantages. It was expected that many respondentswould merely seek to emphasize certain points already covered in the surveyin previous questions, but it was hoped as well that some respondents mightprovide new insights which the author had possibly missed. The questionswere worded as follows. Q3.1 What are the main ADVANTAGES (business, functional,technical, etc.) of a cloud based BPM solution?[Text box] Q3.2 What are the main DISADVANTAGES (business, func-tional, technical, etc.) of a cloud based BPM solution? [Text box]SECTION 4. The nal section of the survey was entitled About you andwas intended to gather relevant personal data relating to the respondents, aswell as allow them to comment on the survey. Q4.1 Which of the below best describes your primary role withrespect to with BPM? business analyst management level user of BPM methods and/or technologies end user of BPM technology software developer 43 54. 3. METHODS researcher or academic researcher or academic student other Q4.2 What is the size of your company by number employees? < 50 50249 250999 10004,999 > 5,000 n/a Q4.3 Which sector does your company primarily operate in? oil and gas, mining, or agriculture manufacturing services IT services n/a Q4.4 Please use the space below to provide any additional re-marks about Cloud BPM and/or to comment on this survey. [Textbox] The nal question was intended to elicit comments from the respondentsabout the form and content of the survey, but also to give respondents a chanceto bring to light any important issues that may have been omitted from thesurvey.44 55. 3. METHODS3.4Proposed denitionThe Review (literature review) and Test (online survey) phases of the projectwere described in the above two sections. However, in terms of the researchprocess, prior to the Test phase the Build phase had to be completed. Thisentailed the process of distilling the results of the literature review into aproposed denition of Cloud BPM and the formulation of the project hy-pothesis. The results of this phase, that is, the proposed denition of CloudBPM, are presented in Chapter 4.3.5Evaluation of proposed denitionThe nal phase of the project, the Reect phase, focused on the comparisonof the proposed denition of Cloud BPM with the results of the online survey.This is discussed in Chapter 5.3.6SummaryIn this chapter the various phases, tasks and research methods used duringthis project have been detailed. The research method followed a sequentialprocess survey, build, test, reect and comprised a literature review andan online expert survey as its pillars, with the proposed denition of CloudBPM and its evaluation representing their respective fruits. In the next chapter, the results of the literature review and the onlinesurvey are presented. Based on the literature survey, a tentative denition ofCloud BPM is proposed, and the results of the online survey are presentedin a summarized form. 45 56. 4. Results4.1 IntroductionThis chapter presents the results of the research conducted with the objec-tive of dening Cloud BPM. First, the denition of Cloud BPM that wasformulated and distilled from the comprehensive literature review (see Chap-ter 2) is presented. Following that are presented the results of the online expertsurvey, which was devised to test the validity of the proposed denition bycomparing it with the views of experts in the eld, as far as these could beinferred from their responses to the survey questions.4.2 Literature analysisHaving completed the literature review, the next phase of the project was toto summarize the key ndings of the literature survey, in anticipation of theformulation of a denition of Cloud BPM. The key ndings were listed (seeAppendix F) and working from these points, a series of assertions regardingCloud BPM was developed. In the next section, a descriptive denition and characterization of CloudBPM is proposed. Following this, the denition is presented as the hypothe-sis, claried and expressed as a series of assertions that will encapsulate thedenition and characteristics of Cloud BPM. It is this series of assertions en-capsulating the hypothesis regarding Cloud BPM that the online survey wasintended to test.4.2.1Description of Cloud BPMDenition. Although BPM is both a methodology and a set of tools, CloudBPM clearly falls into the category of a technology: Cloud BPM is a specic 46 57. 4. RESULTStype of technology which fully supports the practice of BPM, that is, it is aBPMS, the key characteristics of which are its on-demand availability and itspay-per-use pricing model.Market. Research indicates that Cloud BPM is of primary interest to smalland medium size enterprises, as well as for localized pilot BPM projects withinlarge enterprises. In the case of SMBs, the low startup costs associated withCloud BPM platforms mean that there is a lower barrier to entry to this tech-nology. As for large enterprises, low startup costs mean that new departmentalBPM initiatives can be trialled, and in many cases implemented, without theneed to involve the IT department and its cumbersome procurement proce-dures.Functionality. Cloud BPM functionality should support all of the primaryphases of the business process lifecycle (design, congure, enact, diagnose)and be capable of supporting both structured and unstructured processes,and both human task centred and automated processes, or a combination ofthese.Delivery. A Cloud BPM system should be a SaaS or PaaS oering, with amulti-tenant architecture that allows frequent and instant system upgrades.A signicant number of established vendors provide BPMS oerings that aredeployable either in the cloud or on-premise, with identical functionality, ac-cording to customer needs.Advantages.The main advantage of Cloud BPM is the lower startup costsand the exibility aorded by a system that is available as a service. A furtherimportant benet is the elasticity of the service provided, with the associatedcost benets.Disadvantages.The main disadvantages associated with Cloud BPM arethe security concerns that arise from the hosting of valuable business datao-premise. Future trends.Although there is no clear link between added function-ality and cloud architecture at this moment, current trends point towards the47 58. 4. RESULTSgrowing importance of the cloud platform (PaaS) sector of cloud provision,and BPM as a practice and technology is perfectly suited to this development.A BPM platform in the cloud will provide the exibility and agility that iskey to successful business process management, allowing all of the advantagesof a SaaS oering, without its restrictions in terms of functionality and cus-tomizability.4.2.2Hypothesis statementBased on the above ndings of the literature review, a hypothesis is proposedbelow. The hypothesis statement consists of a series of assertions, the asser-tions being of two types. The rst type are rm assertions (Type 1) whichcan be tested by a consideration of the null hypothesis, that is, the negationof the assertion. The second type of assertions are tentative (Type 2), thatis, the assertion is made merely for the purpose of gathering further researchdata in the form of respondents views on a particular aspect of the domain,with a view to deciding the validity of the assertion. Type 1: Cloud BPM is a BPM system that is deployed o-premise and consumed as a service. A cloud BPM system should be capable of executing the core activi- ties that support business process management, i.e. process modelling, process execution, process monitoring, and process analysis. Cloud BPM is essentially a reference to a delivery model for BPM tools and technology, rather than to a new type of functionality. Cloud BPM is a solution which is attractive mainly to the SMB market. New BPM initiatives pursued using a Cloud BPM solution will attract a lower level of business risk over traditional methods, and this charac- teristic is a driver for the adoption of Cloud BPM. Cloud BPM entails signicant security risks. The full benets of Cloud BPM will be more fully realized when the application is purpose built for cloud.48 59. 4. RESULTS The primary advantages of Cloud BPM relate to cost benets and quickerstartup times, and elasticity of service. Type 2: Cloud BPM is suitable for complex processes. The benets of Cloud BPM are not dependent on the full IT stack beingcloud based. Notwithstanding its high strategic value to the organization, BPM issuitable for a cloud implementation. Cloud BPM may have some advantages with respect to its integrationwith social technologies. Taken together, these assertions constitute the hypothesis, that is, theseassertions are considered to be an encapsulation of the term Cloud BPM asis is currently employed by BPM vendors, analysts and practitioners. Theseassertions will be tested by way of an online survey, which will present theassertions in the form of a variety of questions. The structure of the online survey has been mentioned previously, in Sec-tion 3.3. The following section provides a summary of the survey results. Adiscussion of the results and their implications for the proposed hypothesisfollows in Chapter 5.4.3Survey questionnaire resultsAs discussed in the previous chapter (Chapter 3), the online survey question-naire Cloud BPM a survey was posted to 16 dierent BPM groups onLinkedIn and over a period of two weeks received 38 responses. The results ofthe survey are summarized in the sections below.14.3.1 Respondent-specic informationThe survey included three personal questions relating to the respondents their role in relation to BPM, the size of the company they are associatedwith (if any), and the industrial sector of the company. 1 The summary of responses generated by the online survey is shown in Appendix G.49 60. 4. RESULTSThe results of these three questions are given below in Figures 4.1, 4.2 and4.3 and are summarized as follows.BPM role. There was a good mix of respondents with re