Critical success factors in BPM implementation: Creating AI supported decision engine for the business

Master Degree Program in Information Systems and Technologies Management Critical success factors in BPM implementation: Creating AI supported decision engine for the business. Victor Lipov Master Thesis presented as partial requirement for obtaining the Master Degree in Information Management NOVA Information Management School Instituto Superior de Estatística e Gestão de Informação Universidade Nova de Lisboa MGI NOVA Information Management School Instituto Superior de Estatística e Gestão de Informação Universidade Nova de Lisboa Critical success factors in BPM implementation: Creating AI supported decision engine for the business. by Victor Lipov Master Thesis presented as the partial requirement for obtaining a Master's degree in Information Management, specialization in Information Systems and Technologies Management Supervised by Pedro Manuel Maia Malta, PhD NOVA Information Management School July, 2024 i STATEMENT OF INTEGRITY I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Rules of Conduct and Code of Honor from the NOVA Information Management School. [Lisboa, 13/07/2024] ii ABSTRACT In today's rapidly evolving business environment, achieving operational efficiency, transparency, and market responsiveness is crucial for success. Business Process Management (BPM) is a critical tool in this pursuit, encompassing the design, configuration, enactment, and analysis of business processes to drive continuous improvement. Recent advancements in Generative Pre-trained Transformer (GPT) models offer new opportunities to enhance BPM through improved decision-making and progress-tracking capabilities. However, there is a significant gap in the literature concerning the systematic development of AI-powered decision engines to support BPM implementation. This thesis addresses this gap by partially employing Design Science Research Methodology (DSRM) with some limitations described in the document to develop a decision engine artifact tailored for BPM adoption. The research utilizes concept of Critical Success Factors (CSFs) essential for BPM, incorporating AI to automate and optimize decision-making processes. The methodology involves a thorough literature review, the conceptualization and design of a decision engine model, and the development of a prototype. This prototype is evaluated through expert reviews, providing critical insights into its effectiveness and practical applicability. The findings indicate that the integration of AI, particularly GPT models, can significantly enhance BPM by providing dynamic insights and automating complex decision-making processes. The developed decision engine prototype offers a structured approach to BPM adoption, tailored to different organizational stages and supported by a comprehensive list of CSFs. This research contributes to both academic knowledge and practical applications, offering a foundational framework for organizations to enhance their BPM efforts and advance the dialogue on data-driven decision-making in business process implementation. KEYWORDS Business Process Management; Artificial Intelligence; Decision Engine; Critical Success Factors Sustainable Development Goals (SDG): iii TABLE OF CONTENTS Statement of Integrity ...................................................................................................... i Abstract ............................................................................................................................ ii List of Figures .................................................................................................................... v List of Tables .................................................................................................................... vi List of Abbreviations and Acronyms ............................................................................... vii 1. Introduction ................................................................................................................ 1 2. Literature review ........................................................................................................ 2 2.1. Key Concepts and Definitions for AI-related things ............................................. 2 2.1.1. Artificial Intelligence (AI) .............................................................................. 2 2.1.2. Large Language Models (LLMs) ..................................................................... 2 2.1.3. GPT (Generative Pre-trained Transformer) .................................................. 3 2.1.4. Business Process Management (BPM) .......................................................... 3 2.1.5. Critical Success Factors (CSFs) ...................................................................... 3 2.1.6. Decision Engine ............................................................................................. 3 2.1.7. Sentiment Analysis ........................................................................................ 3 2.2. Detailed Explanations of AI Components ............................................................ 3 2.2.1. Machine Learning (ML) ................................................................................. 3 2.2.2. Natural Language Processing (NLP) .............................................................. 4 2.2.3. Deep Learning ............................................................................................... 4 2.3. Identification of current state of bpm adoption of legal entity ........................... 4 2.4. Role of AI in BPM ................................................................................................. 6 3. Methodology .............................................................................................................. 9 3.1. Problem Identification ......................................................................................... 9 3.2. Objectives of a Solution ....................................................................................... 9 3.3. Design and Development ..................................................................................... 9 3.4. Summary of Anticipated Outcomes ................................................................... 10 4. Empirical Study ......................................................................................................... 12 4.1. Approach to decision engine prototype ............................................................ 12 4.2. Decision engine prototype ................................................................................. 13 4.3. Simulated Example of practical application. ...................................................... 16 5. Results and discussion .............................................................................................. 19 5.1. Exposition of Empirical Results .......................................................................... 19 iv 5.2. Validation of Empirical Results .......................................................................... 19 5.2.1. CEO ............................................................................................................. 20 5.2.2. CFO .............................................................................................................. 20 5.2.3. Manager ...................................................................................................... 21 5.2.4. Summary of the feedback: .......................................................................... 21 5.3. Discussion .......................................................................................................... 22 5.4. DE Prototype Implementation Adjustments Based on Expert Suggestions ....... 22 6. Conclusions and future works ................................................................................... 24 6.1. Synthesis of the Work Done with Highlight to the Recommendations Made ... 24 6.1.1. Key Recommendations: .............................................................................. 24 6.2. Limitations of the Work Done ............................................................................ 25 6.3. Further Work Possible to Do .............................................................................. 25 Bibliographical Referenczoes ......................................................................................... 26 Appendix A - Ethics Committee Report ......................................................................... 28 Appendix B – Decision engine prototype ...................................................................... 29 Appendix C - Expert Questionnaire ............................................................................... 30 v LIST OF FIGURES Figure 2.4.1 - BPM lifecycle enriched with AI ............................................................................ 7 Figure 4.1.1 - DE prototype logic ............................................................................................. 12 Figure 4.2.1 - BPM Adoption Stage Progression ...................................................................... 15 Figure 4.2.2 - CSF Scoring Approach ........................................................................................ 15 vi LIST OF TABLES Table 2.3.1 - Top 10 factors affecting BMP implementation (Castro et al., 2019) .................... 5 Table 2.3.2 - CSFs for different stages of BPM adoption (Buh et al., 2015) ............................... 6 Table 4.2.1 - BPM Adoption Stages ......................................................................................... 13 Table 4.2.2 - Critical Success Factors (CSFs) for Stage 1 .......................................................... 14 Table 4.3.1 - BPM adoption simulation ................................................................................... 16 Table 5.4.1 - DE adjustments based on feedback .................................................................... 22 vii LIST OF ABBREVIATIONS AND ACRONYMS AI Artificial Intelligence BPM Business Process Management CSF Critical Success Factors DSRM Design Science Research Methodology GPT Generative Pre-trained Transformer LLM Large Language Models NLP Natural Language Processing ROI Return on Investment SDG Sustainable Development Goals DE Decision Engine 7 domains and tasks (Teubner et al., 2023) and GPT-like models, have significantly enhanced the ability to analyze and identify both explicit and implicit processes within an organization. It means that you can apply GPT-like models to structured and unstructured data to dig the processes or develop solution to the problem. In the research done by (Weinzierl et al., 2024), there are three frequent examples of using ML: • providing decision support through predictions • discovering accurate process models • improving resource allocation. Keeping this in mind, researchers have developed several BPM-specific solutions that extract information from textual documents using Natural Language Processing techniques. (Grohs et al., 2023) In his research, Grohs et al., 2023 illustrated how LLM can be utilized for practical BPM tasks which required textual documents as input. ChatGPT4 was utilized for this task. For all three tasks, GPT4 performs similarly to or better than the benchmark, i.e., specific applications for the respective task (Grohs et al., 2023). There are also increasing number of services to Convert between Process Models and Natural Language Text utilizing graphical modeling languages. It allows stakeholders to collectively describe executable business processes on a conceptual level. (Freytag et al., n.d.). Comprehensive literature review being done on the topic of “Machine learning in business process management” by (Weinzierl et al., 2024). Resulted in enriched BPM lifecycle BPM lifecycle of Dumas et al. (2018). Which gives higher level practical view perspective on the areas, where AI can be applied. Figure 2.4.1 - BPM lifecycle enriched with AI 8 From the perspective of our research, it helps to understand better the ongoing processes in the organization and track the processes which are adopted to BPM. Machine Learning (ML) has been particularly useful in improving certain parts of Business Process Management (BPM). These improvements mostly happen in areas where there is a lot of data, such as discovering new processes, analyzing them, and keeping an eye on them over time. In these areas, data about business activities is usually available in detailed records, which ML can easily use (Sarker, 2021; Weinzierl et al., 2024). However, researchers suggest that we could make even more progress by using techniques from fields like natural language processing (NLP) and computer vision. NLP helps computers understand and work with human language, while computer vision helps them understand images and videos. By combining these techniques with ML, we could create new and better ways to manage business processes that go beyond just using data logs. This could lead to innovative solutions that are more effective and versatile. Weinzierl et al., 2024 also propose to consider the enterprise process network instead of isolated business processes. In doing that, ML applications receive input data from multiple data sources, including control-flow information from different business processes and process context information related to the business processes. Leveraging LLMs and GPT models facilitates comprehensive process identification, dynamic roadmap development, real-time implementation monitoring, and the tracking of qualitative metrics. These capabilities address significant gaps in traditional BPM approaches, paving the way for more effective and efficient systems. By continuously integrating AI-driven insights, organizations can ensure their BPM initiatives are not only successful but also adaptable to future challenges and opportunities (Grohs et al., 2023; Weinzierl et al., 2024). In addition to traditional performance metrics, tracking qualitative metrics such as sentiment and involvement (how people feel and how involved they are in their work) is vital for comprehensive BPM evaluation. AI technologies, including sentiment analysis and engagement tracking tools, can analyze qualitative data to provide deeper insights into employee morale, customer satisfaction, and overall organizational engagement. These insights provide a deeper understanding of the workplace atmosphere and customer experiences, allowing organizations to pinpoint areas that need improvement and take specific actions to enhance overall BPM results. 9 3. METHODOLOGY This section outlines the methodology adopted for this research, aiming to identify and utilize Critical Success Factors (CSFs) within BPM to develop a decision engine artifact that supports BPM implementation. The initial idea was to apply Design Science Research Methodology (DSRM) as a framework for this study (Hasan & Sharifzadeh, 2020; Hunziker & Blankenagel, 2024). However, due to the lack of access to a real-world company, it was not feasible to accomplish stages of DSRM such as Demonstration and Evaluation fully. Nevertheless, some parts of DSRM were applied, such as Problem Identification, Objectives of a Solution, Design and Development. Detailed description provided below: 3.1. PROBLEM IDENTIFICATION Objective: The primary goal at this stage is to identify the current challenges and gaps in BPM implementation. This involves a deep dive into existing literature and research to pinpoint areas where BPM practices are lacking, particularly focusing on the systematic development of AI-supported decision engines. The objective is to establish a clear understanding of the problems faced by organizations in adopting and optimizing BPM systems. Approach: A comprehensive literature review was conducted to gather insights into the current state of BPM adoption, common challenges, and the potential role of AI in enhancing BPM. This involved analyzing various scholarly articles, industry reports, and case studies to identify recurring themes and issues. The literature review also included an examination of existing AI technologies and their applications in BPM, providing a foundational understanding of how AI can be leveraged to address identified challenges. At this stage we identified importance of CSFs for BPM adoption as well as breakdown of BPM adoption to different Stages (states of organization). 3.2. OBJECTIVES OF A SOLUTION Objective: The next step is to define the goals and requirements for the AI-supported decision engine. This involves setting clear, measurable objectives that the decision engine aims to achieve. The objective is to ensure that the proposed solution addresses the identified challenges and gaps in BPM implementation effectively. Approach: Insights from the literature review and expert interviews were utilized to outline the objectives. The objectives were refined through iterative feedback, ensuring they are aligned with the needs of organizations and the capabilities of AI technologies. This approach ensured that the solution is both theoretically sound and practically relevant. 3.3. DESIGN AND DEVELOPMENT Objective: The primary goal here is to develop the decision engine artifact based on the identified CSFs and the defined objectives. This involves a detailed design and 10 development process, ensuring that the decision engine is robust, scalable, and capable of addressing the identified challenges in BPM implementation. Stage 1: Identification of Current State of BPM Adoption Approach: The current BPM adoption stage of the organization was assessed using a structured framework. This framework was derived from the research of Buh et al. (2015), which categorizes BPM adoption into various stages and identifies associated CSFs for each stage. Stage 2: Identify CSFs for Respective State Approach: The list of critical success factors (CSFs) was condensed from Buh et al. (2015) research to focus on the most relevant factors for each BPM adoption stage. This involved prioritizing CSFs that have the highest impact on BPM success and are most applicable to the organization's current state. This process included: • Reviewing the detailed descriptions and justifications of each CSF. • Evaluating their applicability based on the organization's specific context and needs. • Engaging with subject matter experts to validate the selection and prioritization of CSFs. Stage 3: Develop a Scoring Approach for Respective State Approach: The scoring approach was developed with the support of GPT-4 through a series of prompts and practical experience. This process involved: • Defining clear criteria for each CSF based on literature and expert input. • Developing a rating scale (e.g., 1 to 5) for each criterion to quantify the level of achievement or performance. • Testing the scoring approach through simulated scenarios and expert validation to ensure its reliability and validity. Prompts example: • Please suggest list of criteria which can be used for scoring {proposed scoring approach}. • What kind of scales can be applied to {measurement criteria}. Responses were critically analyzed and adopted to the topic. Stage 4: Critical Evaluation Approach: Interviews with experts across different levels of the corporate hierarchy were conducted to evaluate the prototype from various perspectives. This included: • Conducting structured interviews with CEO level, CFO level and line managers to gather diverse insights. • Analyzing the feedback to identify common themes, strengths, and areas for improvement in the decision engine. 3.4. SUMMARY OF ANTICIPATED OUTCOMES The anticipated outcomes of this research include: 11 • A list of Critical Success Factors (CSFs) for BPM implementation, tailored to different stages of BPM adoption. • A theoretical model for an AI-supported decision engine that incorporates these CSFs, providing a structured approach to enhance BPM practices. • Critical insights from experts across different levels of expertise in BPM. This research aspires to contribute to both academic knowledge and practical applications in BPM. By providing a foundational framework for AI-enhanced BPM implementation, it aims to support organizations in achieving higher levels of operational efficiency, transparency, and market responsiveness. 12 4. EMPIRICAL STUDY 4.1. APPROACH TO DECISION ENGINE PROTOTYPE Development of Decision Engine (DE) can address complexity and add transparency to the process of BPM adoption. As per definition, decision engines are sophisticated software systems designed to automate the decision-making processes within an organization. It operates by analyzing input data through a set of predefined rules, algorithms, or machine learning models to make decisions or recommendations without human intervention (purely based on available data). These engines are particularly valuable in scenarios where fast, consistent, and accurate decisions are crucial, such as in credit scoring, fraud detection, customer service, and many other operational processes. For this paper, we decided step aside from generalized factors and align decision engine with different stages of BPM adoption. To do this, we developed the prototype in the following logic: Figure 4.1.1 - DE prototype logic The idea behind is that before we start, we should understand where we are now. When we understand where we are, we can decide what to do next and how to do it. 13 4.2. DECISION ENGINE PROTOTYPE From our perspective decision engine prototype should be adoptable and practical. It should guide the end user through the different stages; therefore, it should help answering two initial questions: • Where the organization is now? • What should it do next? To answer to the first question the prototype includes description of BPM adoption stages as per Buh et al., 2015. Identification, and Moving forward actions were derived with support of ChatGPT4 using prompts like: • based on your experience, how can organization be identified at {Stage} • please suggest possible actions for the organization to advance further and summarized to the table below: Table 4.2.1 - BPM Adoption Stages Stage Description Identification Moving Forward Actions Awareness and Understanding of BPM Organizations recognize the importance of BPM. Process improvement is ad hoc, reactive, and unstructured. BPM is a new concept or there is limited understanding of process management principles organization-wide. Introduce formal BPM training and establish a sense of urgency for process improvement. Desire to Adopt BPM Organizations are motivated to adopt BPM and are beginning to plan their initiatives. Management shows interest in BPM, and initial resources are allocated. Secure top management support, appoint a project champion, and define business drivers. BPM Projects Organizations execute specific BPM projects with defined objectives and plans. Clear objectives, plans, and external consultant guidance are in place. Communicate objectives, seek professional guidance, and motivate team members. BPM Program BPM is integrated into the organization’s broader strategy and programs. Continuous top management support, professional guidance, and effective communication are evident. Ensure ongoing support, maintain consultant partnerships, and identify KPIs. Productization of BPM BPM practices are standardized and Key performance indicators (KPIs) are Focus on training, securing top 14 productized within the organization. defined, employees are educated and motivated, and BPM is embedded in decision-making. management endorsement, and refining processes based on KPIs. For example: Stage 1: Awareness and Understanding of BPM • Description: Organizations at this stage have begun to recognize the importance of BPM. Process improvement is typically ad hoc, reactive, and unstructured. The organization recognizes BPM as a potential value-add and begins to understand its principles and benefits. • Identification: If BPM is a new concept or if there is limited understanding of process management principles organization-wide, it's likely at this stage. • Moving Forward: Introduce formal BPM training and establish a sense of urgency for process improvement. If the description matches the type of organization, the organization can become aware of focus points and suggestions relevant for the stage. In our example: • Focus: Education on BPM importance and its impact on performance. • Education and Communication: Conduct workshops and training sessions to educate stakeholders about the benefits and principles of BPM. • Visible Executive Support: Secure and showcase executive endorsement to highlight the importance of BPM within the organization. • Case Studies and Success Stories: Share case studies of successful BPM implementations to illustrate potential gains and encourage buy-in. The same applies to potential problems. As a result, the organization gets a list of CSFs to consider. In our example, for the first stage of BPM adoption, CSFs include empowerment of employees, customer focus, and openness to changes (Buh et al., 2015). Table 4.2.2 - Critical Success Factors (CSFs) for Stage 1 CSF Focus Points Actions Empowerment of Employees Training and workshops to enhance understanding of BPM. Conduct monthly BPM training sessions and interactive workshops. 15 Customer Focus Implement feedback loops to align processes with customer needs. Introduce surveys and feedback forms post-customer interactions. Openness to Changes Encourage a culture of adaptability and continuous improvement. Develop and communicate a structured change management plan, implement a recognition program for successful changes. By knowing CSFs, the organization can develop scoring approaches, measurement criteria, scales, etc. Below is a flowchart representing the progression through different stages of BPM adoption, highlighting key actions and focus points at each stage. Figure 4.2.1 - BPM Adoption Stage Progression A radar chart showing the scoring of different CSFs helps visualize the strengths and areas for improvement for each factor. Figure 4.2.2 - CSF Scoring Approach Including these tables and charts enhances the clarity and practicality of the prototype, making it easier for organizations to navigate through the stages of BPM adoption. 16 4.3. SIMULATED EXAMPLE OF PRACTICAL APPLICATION. With the support of ChatGPT4 we have done series of practical simulations on how it can be applied to the real-world scenarios. Company Profile: Name: Tech Innovators Inc. Industry: Technology Solutions Current BPM Adoption Stage: Awareness and Understanding of BPM The company decides to start BPM adoption smoothly by focusing on two key processes: the Customer Support Process and the Product Development Process. Prompts used: • Please generate list of possible tasks for {step} of {action / task} of {stage} • What possible results/ outcomes can be for {step} of {action / task} of {stage} Table 4.3.1 - BPM adoption simulation Step Action/Task Expected Result 1. Initial Assessment Surveys Conduct internal surveys to gather insights from employees. Clear understanding of current BPM knowledge and engagement. Interviews Conduct interviews with key stakeholders. Identify pain points and opportunities for improvement. Document Review Analyze existing process documentation, training records, and performance metrics. Identify gaps and areas needing enhancement. Outcome The assessment revealed that while there was an acknowledgment of the need for BPM, the process improvement initiatives were ad hoc, reactive, and unstructured. This placed Tech Innovators in the "Awareness and Understanding of BPM" stage. 2. Define CSFs Empowerment of Employees Enhance employees' understanding of BPM and encourage contributions. Improved employee engagement and participation in BPM initiatives. Customer Focus Implement mechanisms to gather and act on customer feedback. Better alignment of customer needs with support processes. 23 customer feedback, and foster adaptability. and increased flexibility. clear financial metrics. financial clarity, and adaptability. Develop Scoring Approach Develop criteria and rating scales for each CSF based on literature and expert input. Structured evaluation of BPM effectiveness. Manager: Ensure scoring is simple and actionable. Practical and easyto-apply scoring system for evaluating BPM effectiveness. Implement Decision Engine Conduct BPM training sessions, introduce feedback mechanisms, and communicate change management plans. Increased understanding of BPM, enhanced customer feedback integration, and higher adaptability to change. CEO: Balance AI insights with human interaction; CFO: Ensure costbenefit analysis; Manager: Support career development. Improved BPM understanding, better feedback integration, costeffective decisions, and enhanced employee growth. Evaluation and Results Measure improvements in employee engagement, customer satisfaction, and adaptability. Demonstrated improvements in BPM practices and stakeholder engagement. CFO: Include financial KPIs; Manager: Regular feedback sessions. Comprehensive evaluation showing enhanced BPM practices, financial performance, and stakeholder engagement. This research contributes to both academic discourse and practical applications in BPM. By developing and validating an AI-supported decision engine prototype, it offers a foundational framework for organizations to enhance their BPM efforts. The framework can be used in any type of organization and on different stages of development, but practical application should be supported by further researches. The integration of AI technologies and human-centric design principles ensures that BPM initiatives are effective, adaptable, and aligned with organizational goals. Future research could focus on refining the decision engine based on broader industry feedback and exploring its application in different organizational settings. 24 6. CONCLUSIONS AND FUTURE WORKS In this final chapter, we synthesize the work done in this research, highlight the recommendations made, discuss the limitations encountered, and suggest possible directions for further research. 6.1. SYNTHESIS OF THE WORK DONE WITH HIGHLIGHT TO THE RECOMMENDATIONS MADE This thesis explored the integration of AI-supported decision engines into Business Process Management (BPM) to enhance BPM adoption and implementation. The primary contributions of this research are: Literature Review: A comprehensive literature review identified Critical Success Factors (CSFs) crucial for BPM implementation and examined the role of AI, particularly Large Language Models (LLMs) like GPT, in BPM. Development of a Decision Engine Prototype: Using some of the components of Design Science Research Methodology, we developed a decision engine prototype tailored to different stages of BPM adoption. The engine was designed to guide organizations through awareness, adoption, project execution, and productization stages, providing stagespecific recommendations. Empirical Study: The prototype was tested through a simulated case study of Tech Innovators Inc., demonstrating its practical application and effectiveness in improving BPM practices. Expert Validation: Interviews with industry experts validated the approach and provided critical feedback for improvement. The recommendations from experts included managing expectations, incorporating financial metrics, ensuring clear communication, and balancing AI-driven insights with human judgment. 6.1.1. Key Recommendations: Realistic Communication: Clearly communicate the benefits and limitations of BPM to avoid overhyping and to build a solid foundation. Leadership Commitment: Ensure strong and empathetic leadership to drive BPM initiatives. Financial Metrics: Integrate financial performance metrics to demonstrate ROI and secure executive support. Human Interaction: Balance AI insights with human-centric features to build trust and facilitate engagement. 25 Training and Education: Conduct practical training and workshops to enhance employee understanding. 6.2. LIMITATIONS OF THE WORK DONE Several limitations were encountered during this research: Lack of Real-World Company Access: The inability to access a real-world company limited the scope of empirical validation. The simulated case study provided valuable insights but lacked the complexity of real-world implementation. Scope of Expert Interviews: While expert feedback was crucial, the sample size was small. A broader range of industry perspectives could provide more comprehensive validation. Prototype Testing: The decision engine prototype was tested in a controlled environment. Real-world testing might reveal additional challenges and areas for refinement. 6.3. FURTHER WORK POSSIBLE TO DO Future research could address the limitations and expand on the findings of this thesis: Real-World Implementation: Conduct case studies with real-world companies to validate and refine the decision engine prototype in diverse organizational settings. Broader Expert Feedback: Engage with a larger and more diverse group of industry experts to obtain comprehensive validation and insights. Advanced AI Integration: Explore the integration of more advanced AI techniques, such as reinforcement learning and hybrid AI models, to enhance the decision engine's capabilities. Longitudinal Studies: Conduct longitudinal studies to assess the long-term impact of AIsupported decision engines on BPM adoption and performance. Customization for Different Industries: Develop industry-specific versions of the decision engine to address unique challenges and requirements in sectors such as healthcare, finance, and manufacturing. By addressing the limitations and pursuing further research, we can continue to advance the field and support organizations in achieving greater operational efficiency and market responsiveness. 26 BIBLIOGRAPHICAL REFERENCZOES Buh, B., Kovačič, A., & Indihar Štemberger, M. (2015). Critical success factors for different stages of business process management adoption – a case study. Economic ResearchEkonomska Istraživanja, 28(1), 243–258. https://doi.org/10.1080/1331677X.2015.1041776 Castro, B. K. D. A., Dresch, A., & Veit, D. R. (2019). Key critical success factors of BPM implementation: A theoretical and practical view. Business Process Management Journal, 26(1), 239–256. https://doi.org/10.1108/BPMJ-09-2018-0272 Dumas, M., La Rosa, M., Mendling, J., & Reijers, H. A. (2018). Fundamentals of Business Process Management. Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-662-565094 Freytag, T., Kanzler, B., Leger, N., & Semling, D. (n.d.). NLP as a Service: An API to Convert between Process Models and Natural Language Text. Grohs, M., Abb, L., Elsayed, N., & Rehse, J.-R. (2023). Large Language Models can accomplish Business Process Management Tasks (arXiv:2307.09923). arXiv. http://arxiv.org/abs/2307.09923 Han, J., Pei, J., & Tong, H. (2022). Data mining: Concepts and techniques. Morgan kaufmann. Hasan, R., & Sharifzadeh, R. (2020). Design Science Research Methodology As a SolutionOriented Methodology. Hunziker, S., & Blankenagel, M. (2024). Design Science Research Design (pp. 97–116). https://doi.org/10.1007/978-3-658-42739-9_6 Khurana, D., Koli, A., Khatter, K., & Singh, S. (2023). Natural language processing: State of the art, current trends and challenges. Multimedia Tools and Applications, 82(3), 3713– 3744. https://doi.org/10.1007/s11042-022-13428-4 Liu, B. (2020). Introduction. In Sentiment Analysis: Mining Opinions, Sentiments, and Emotions (pp. 1–17). Cambridge University Press. Sarker, I. H. (2021). Machine Learning: Algorithms, Real-World Applications and Research Directions. SN Computer Science, 2(3), 160. https://doi.org/10.1007/s42979-02100592-x Teubner, T., Flath, C. M., Weinhardt, C., Van Der Aalst, W., & Hinz, O. (2023). Welcome to the Era of ChatGPT et al.: The Prospects of Large Language Models. Business & Information Systems Engineering, 65(2), 95–101. https://doi.org/10.1007/s12599-023-00795-x van der Aa, H., Carmona, J., Leopold, H., Mendling, J., & Padró, L. (2018, August). Challenges and Opportunities of Applying Natural Language Processing in Business Process Management. Van Der Aalst, W. M. P. (2013). Business Process Management: A Comprehensive Survey. ISRN Software Engineering, 2013, 1–37. https://doi.org/10.1155/2013/507984 Weinzierl, S., Zilker, S., Dunzer, S., & Matzner, M. (2024). Machine learning in business process management: A systematic literature review. Expert Systems with Applications, 253, 124181. https://doi.org/10.1016/j.eswa.2024.124181 27 Weske, M. (2012). Introduction. In M. Weske (Ed.), Business Process Management: Concepts, Languages, Architectures (pp. 3–23). Springer Berlin Heidelberg. https://doi.org/10.1007/978-3-642-28616-2_1 Copeland, B. (2024). artificial intelligence. Encyclopedia Britannica. https://www.britannica.com/technology/artificial-intelligence McDonough, M. (2024). large language model. Encyclopedia Britannica. ChatGPT4. (2024). ChatGPT. https://chatgpt.com/ Amazon Web Services. (2024). What is GPT? Amazon. https://aws.amazon.com/what-is/gpt/ 28 APPENDIX A - ETHICS COMMITTEE REPORT 29 APPENDIX B – DECISION ENGINE PROTOTYPE Stage of BPM adoption Focus points and suggestions relevant for the stage Potential problems at the respective stage CSFs to address and methods to use Proposed Scoring approach Measurment criteria Scale Data sources Example of insight and recommendations Decision logic Empowerment of Employees: Offer training and workshops to improve employees' understanding of BPM and its benefits, and create a suggestion system that enables them to contribute ideas. Emplo yee Empowerment: Scoring could be based on employee training initiatives, the presence of a suggestions program, and the level of autonomy employees have in their roles. Number of BPM-related training sessions held. Employee participation rate in BPM training and workshops. Scale: 1 (No involvement) to 5 (Full involvement and empowerment in BPM activities). Internal suggestion systems, HR records. Actionable Insight: Employees are not suggesting process improvements, indicating a lack of empowerment. Recommendation: Develop and implement an idea management system to capture employee suggestions. Hold regular workshops and training sessions to educate employees on BPM principles. Encourage managers to foster a culture that values input from all team members. If the average score for employee empowerment is below 3, the decision engine recommends prioritizing employee engagement initiatives. Customer Focus: Implement a feedback loop with customers to understand their needs and expectations, and integrate this feedback into process improvement plans. Customer Focus: Measure the extent to which customer feedback influences process changes and the alignment of processes to customer needs. Existence and utilization of customer feedback mechanisms in processes. Number of process improvements driven by customer feedback. 1 (Not integrated) to 5 (Fully integrated and driving BPM initiatives). Customer service feedback, process documentation. If customer focus is rated high but change openness is low, suggest strategies for more responsive process improvements based on customer feedback. Openness to Changes: Encourage a culture that values adaptability through regular communication about the benefits of change and recognition of teams that embrace and lead successful change initiatives. Openness to Change: Assess the organization's history and readiness for change, the existence of change management strategies, and flexibility in current processes. Existence and utilization of customer feedback mechanisms in processes. Number of process improvements driven by customer feedback. 1 (Resistant to any change) to 5 (Actively embraces change). Chang e log s, project management tools. Involvement and Full Support of Top Management: Ensure that top management not only endorses but is also actively involved in BPM initiatives by leading project kick-offs and regularly reviewing progress. Top Management Invo lvement: Score based on visible support from leadership, allocation of resources for BPM, and communication about the importance of BPM from top management. Frequency and consistency of communication from top management regarding BPM. Investment in BPM initiatives (budget, resources). 1 (No support) to 5 (Highly visible and committed support). Meeting minutes, executive communications. Actionable Insight: Top management involvement in BPM initiativ es is minimal. Recommendation: Schedule regular BPM briefings for top management to report on progress and value. Appoint a senior executive as a BPM sponsor to increase visibility and commitment. Create a BPM steering committee including top management to oversee initiatives. If top management support scores less than 3, trigger a recommendation for executive workshops on BPM value. If a project champion is not identified or scores low, recommend establishing a clear champion role with defined responsibilities and authority. Project Champion: Appoint a respected leader within the organization to act as a BPM champion who will advocate for and drive BPM efforts. Project Champion Presence: Evaluate the appointment of project champions for BPM initiativ es and the authority and resources given to them. Identification and activity level of BPM champions or process owners. Impact assessment of project champions on BPM projects (progress, success stories). 1 (No champion) to 5 (Strong, effective champion promoting BPM). Project documentation, stakeholder surveys. Business Drivers (a Sense of Urgency): Identify and communicate the business imperatives that necessitate BPM, such as market pressure or efficiency gains, to create a sense of urgency for change. Sense of Urgency/Business Drivers: Determine whether there's a clear understanding across the organization of the need for BPM, driven by competitive pressure, efficiency goals, or other business imperatives. Documentation and communication of business imperatives driving BPM adoption. Response time to critical process-related issues or market changes. 1 (No drivers identified) to 5 (Clear drivers identified and communicated with urgency). Strategic plans, internal communications. Clearly Defined Objectives and Plan: Develop a formal project charter for each BPM project that outlines its objectives, scope, expected benefits, and timeline. Project Communication: Score the clarity and frequency of communication about BPM project goals, progress, and outcomes. Clarity and dissemination of project goals and milestones across the organization. Number of documented and communicated BPM project plans. 1 (Very unclear) to 5 (Extremely clear and well understood by all stakeholders). Project charters, communication records. Actionable Insight: Project objectives and plans are not well-communicated across the organization. Recommendation: Utilize a centralized communication platform to disseminate BPM project updates. Include clear communication plans in all BPM project charters. Establish regular all-hands meetings to discuss BPM projects and invite feedback. When scores for project communication and clarity are low, the logic emphasizes improving project charters and communication plans. If motivation to change is high but guidance from consultants is low, the logic suggests a review of consultant selection criteria and engagement effectiveness. Professional Guidance of External Consultants: Engage with consultants who have a proven track record in BPM to provide expertise and an external perspective. External Consultants' Guidance: Rate the extent and effectiveness of external consultancy use, including the transfer of knowledge and practices to internal teams. Consultant engagement score based on project outcomes and feedback. ROI from con sultant-led initiatives versus internally led. 1 (Not involved/no contribution) to 5 (Significant involvement/contrib ution). Consultant performance evaluations, project outcome analyses. People Who Are Willing and Motivated to Change: Select project team members who are open to change and have demonstrated adaptability; involve them in decision-making to keep them motivated. Change Willingness: Measure the organization's readiness for change specific to BPM projects, including staff engagement and participation. Pre and post-training assessment scores on BPM readiness. Employee turnover rates in departments undergoing BPM projects. 1 (High resistance) to 5 (Eager willingness and motivation to change). Engagement surveys, participation records in BPM initiatives. Miscommunication: If not effectively communicated, BPM can be seen as a buzzword rather than a valuable methodology, leading to skepticism. Insufficient Executive Support: With out strong leadership endorsement, BPM efforts may not be taken seriously or prioritized. Overwhelm with Information: An abundance of case studies and information can lead to confusion about where to start or unrealistic expectations. Cultural Resistance: Even with awareness, there can be inherent resistance to change within the organization's culture that is hard to overcome. Misalignment of Expectations: Stakeholders may have different expectations from BPM, leading to conflicts or disillusionment if not managed well. Selection of Incorrect Projects: Choosing the wrong projects for initial BPM efforts can result in a lack of visible success and diminished credibility. Project Management Issues: With out strong pro ject management, BPM projects may face delays, budget overruns, or scope creep. Inadequate Measurement: Poorly defined KPIs or a lack of measurement tools can lead to ineffective monitoring and the inability to demonstrate value. Desire to Adopt BPM Description: The organization acknowledges the need for BPM and is motivated to start adopting BPM practices. There might be some informal process improvements, but no organized effort yet. The organization decides to commit to BPM practices and starts to build internal desire for change and process improvement. Identification: If there is enthusiasm about BPM and discussions about its potential benefits are occurring, but no structured approach has been taken, this indicates stage 2. Moving Forward: Develop a clear vision and strategy for BPM and start engaging stakeholders to support BPM initiativ es. Focus: Building internal buy-in and preparing the organization for change. Culture and Change Management: Foster a culture receptive to change by involving employees in the BPM discussion early on. Address resistance proactively. Stakeholder Engagement: Identify and engage key stakeholders to create a coalition that desires BPM adoption, ensuring their needs and concerns are addressed. BPM Projects Description: The organization has initiated specific BPM projects, focusing on improving particular processes. These are often pilot projects that serve as proof of concept.Specific BPM projects are initiated, often as pilots or initial efforts to apply BPM principles to particular business processes. Identification: If there are one or more BPM projects underway with dedicated resources and goals, the organization is in this stage. Moving Forward: Analyze the outcomes of these projects to refine the BPM approach and prepare for broader implementation. Focus: Gaining experience in BPM and demonstrating early wins. Select Pilot Projects: Choose initial BPM projects carefully, targeting high-impact areas where quick wins are achievable. Project Management Best Practices: Utilize structured methodologies and robust project management practices to ensure the success of BPM projects. Monitoring and Measurement: Implement KPIs to measure the performance of BPM projects, using the data to drive decisionmaking and continuous improvement. Awareness and Understanding of BPM Description: Organizations at this stage have begun to recognize the importance of BPM. Process improvement is typically ad hoc, reactive, and unstructured. Organization recognizes BPM as a potential value-add and begins to understand its principles and benefits. Identification: If BPM is a new concept or if there is limited understanding of process management principles organizationwide, it's likely at this stage. Moving Forward: Introduce formal BPM training and establish a sense of urgency for process improvement. Focus: Education on BPM importance and its impact on performance. Education a nd Co mmunication: Conduct workshops and training sessions to educate stakeholders about the benefits and principles of BPM. Visible Executive Support: Secure and showcase executive endorsement to highlight the importance of BPM within the organization. Case Studies and Success Stories: Share case studies of successful BPM implementations to illustrate potential gains and encourage buy-in. Stage #1 Stage #2 Stage #3 30 APPENDIX C - EXPERT QUESTIONNAIRE 1. In your experience, what are the key elements to focus on during the BPM adoption? 2. How would you describe the most effective methods to introduce BPM concepts to the organization? 3. What role do you see for leadership in BPM adoption? 4. Do you expect any challenges during implementation of BPM? How can they be mitigated? 5. What are the most critical success factors for BPM adoption that you believe should be prioritized? 6. If you can apply a decision engine model for BPM, what features do you think would help you the most? What to avoid and how to structure better? 7. How do you view the potential role of AI in BPM, especially in identifying and redesigning business processes? 8. From your perspective, what types of data sources can be used for the decision engine? Would external inputs like customer feedback add value? 9. Should a BPM tool balance data-driven insights with human judgment? Or be purely technical? 31