Improvement and digitalization of business processes in small-medium enterprises Matas Dumčius Tomas Skersys Department of Information Systems, Faculty of Informatics Department of Information Systems, Faculty of Informatics Kaunas University of Technology Kaunas University of Technology Studentų 50, Kaunas, Lithuania Studentų 50-313a, Kaunas, Lithuania matas.dumcius@ktu.edu tomas.skersys@ktu.lt Abstract — As small-medium enterprises (SMEs) produce engagement and agility [4], [2]. Arguably, the ultimate goal of huge amounts of added value in the market, it is important to BPM is a digitalized enterprise running on optimal business ensure that their business processes are optimized. Business processes throughout its whole life cycle. Even though Process Management (BPM) methodology is popular in large traditionally most of the successful BPM initiatives are enterprises, however, there is lack of information on how it can associated with very large organizations, it has no restrictions be adapted by SMEs considering the financial and on being implemented within the working environment of informational constraints they face. In this research, a set of SMEs whatsoever. However, considering the financial and qualitative and quantitative analysis methods for improving the informational environment in which SMEs are working, the quality of business process models are presented and applied to effectiveness of this methodology might be unpredictable, and optimize the order management business process of a small Lithuanian optical retail business company. Further, a business the results might be even detrimental in case of process management system (BPMS) is presented to digitalize straightforward BPM application in a small-medium sized and support the execution of redesigned process. A brief organization. In order to get positive outcome of BPM discussion of applying process quality analysis methods within initiative in SME, every step in BPM lifecycle should be the selected business domain is presented. It is expected that our analyzed in the context of such business. research will make BPM initiatives more feasible for businesses The main task of this research is to evaluate different types of similar type. of qualitative process evaluation and optimization methods of Keywords— business process improvement, business process small optical retail business order management process. Then management, business process digitalization, business process test the redesigned processes by using business process management system, small-medium enterprise, BPMS, BPMN, simulation tool in order to get the best process optimization BPM, quality management. result. Such evaluation and example of process optimization might help to apply BPM methodology in SMEs process I. INTRODUCTION redesign phase easier for other businesses of the same type. Small-medium enterprises are defined as companies with II. BASIC DEFINITIONS AND RELATED WORK staff count of up to 250 employees or turnover lover than 50 million € or balance sheet total less than 43 million € by the A. Definition of Business Process Management European Commission. Such businesses produce vast Business Process Management does not have one specific majority of added value in the economics of European Union. definition. Generally, it is a methodology which aims at As stated in the annual report on European SMEs 2017/2018, improving organization performance by concentrating on those kind of businesses accounted for 47% of the increase in business processes. BPM combines the business management, the value added and 52% of the cumulative increase in quality control and information technologies traditions. The employment in non-financial sector since the global financial origin and composition of BPM can be seen in Fig. 1. crisis [1]. Most of the SMEs are working in the state of very limited resources, accessibility to information and experience in process management and automation. But these fields are of the highest importance when considering that the main goal of business is related to creating value through activity while utilizing resources in the most efficient way. Optimization and documentation of SMEs processes are important as efficient and effective process delivery is a key to a long-term success, business competitiveness, growth and viability [2],[3]. One of the most common ways of improving business processes Fig. 1. Approaches of business process change [5] running in one or across many enterprises is by implementing a Business Process Management (BPM) methodology. BPM initiatives are carried out by following a predefined set of stages – a BPM lifecycle (Fig. 2). At first – a business BPM is a widely spread methodology, which concentrates problem is defined and a target process is selected. Then the on business processes and aims to improve their quality, current state of the process is documented (usually in a form efficiency, compliance, customer integration, employee of process model). When a process model is present qualitative and quantitative analysis are carried out so that any © 2019 for this paper by its authors. Use permitted under Creative improvement could be measured. Then the process is Commons License Attribution 4.0 International (CC BY 4.0) redesigned according to the initially stated issues and a new 22 process model is created. This process model serves a basis behind various incidents or defects [13]. Though this method for the next stage – process implementation. The changes are might be adapted for thorough problem identification and realized and the process is moved to the new state so that analysis in business process model with an intention of performance goals could be achieved. Usually management optimizing it [14]. First step of such analysis is defining changes and process automation via development of perspectives in which stated problems will be explored. 6M information systems are implemented. Once changes are model (machine, method, material, man, measurement, completed, process execution data is collected, analyzed, new milieu), some perspectives from Six Sigma methodology [15] problems or goals are defined and respective actions are taken. or any other model might be applied. Then each stated problem is analyzed in every perspective while searching for the root cause. This type of analysis does not require a full process model, but the results of it might help redesign ongoing process. Impact assessment and issue documentation is type of analysis that typically follows up root cause analysis. As mentioned before, root cause analysis defines problems and their cause. But it does not point out their impact to the whole process, so there is no formal way to prioritize them. One suggestion is to create issue register, in which each problem would have an impact assessment (qualitative or quantitative) such as impact on time, finances or any other metric [6]. If metrics are defined, we can conduct Pareto analysis. In practice Pareto analysis makes an assumption that 20% of problems make 80% of impact [16]. Of course, if there are only few problems stated in the issue register, this type of Fig. 2. BPM lifecycle [6] analysis is unnecessary. All mentioned types of qualitative analysis help process B. Quality of the process analytics to identify existing problems and redesign process Quality is one of the most important parts through the models. But these methods do not specify any information on whole BPM lifecycle. In this case we are talking not about the how process model redesign could be evaluated, or process results of the process – quality of deliverables such as products models could be compared. In such cases quantitative analysis or services – but about execution properties of the process. must be performed to get the required data. Quality models are defined as easily understandable and changeable models with little to no design errors. A modeled Quantitative analysis may be conducted in three different process reveals only activities and decisions, but no categories: analysis of process model metrics which are information can be retrieved directly about its quality metrics derived from software engineering, theoretical process [12]. For this reason, there are numerous qualitative and execution analysis and process simulation. quantitative process analysis methods which help to find Software quality quantitative metrics can be applied in bottlenecks of the process, sources of arising problems, business process model evaluation because of huge similarity compare different process models and come up with an between processes and software – they both process data, have optimized process model. But there is small amount of a structure and are based on a static model [12], [17]. information about application of these methods in the scope of Calculating such metrics as model coupling, cohesion, small-medium sized business. complexity, modularity or size could give good indications C. Process model optmization methods about the quality of a process model. Process model metrics are valuable for process analysts, but process stakeholders are Qualitative process analysis mostly helps to identify usually interested in execution metrics such as execution redundant or weak parts of the process. In the next section we price, duration, quality of results and model flexibility. All will summarize some methods of qualitative process analysis. these parameters can be calculated by performing Flow Value-Added Analysis – the main goal of such analysis is analysis or applying Queuing theory [6]. These methods can to remove non-value adding tasks from the process model. give valuable data, but they are hardly applicable in real world The process is decomposed to the simplest tasks requiring one process models. Flow analysis can be conducted only when action of one process participant. Then those tasks are process models are simple, have only exclusive or parallel assigned one of three categories: gateways and Queuing theory can only give data for one activity. Also Queuing theory calculations are complicated  Value-adding – task that contributes to the final product even for simple situations. For those reasons the most practical or service, solution is process model simulation. The process model  Business value-adding – task that is necessary for simulation can be done by using various tools such as IBM business to be running, WebSphere Business Modeler, ITP Commerce Process Modeler for Visio, ProSim or an open-source solution – BIMP  Non-Value adding – all remaining tasks. simulator [20]. Process simulation software instantiates huge amount of hypothetic process instances and records properties By removing the non-value adding tasks we should be of each execution. Only process execution data, such as utilizing resources in a more efficient way. probabilities of various decisions and duration of activities are Root Cause Analysis – this type of analysis is mostly required as input parameters. This type of qualitative analysis conducted in manufacturing companies to find out the reason allows process analysts to easily compare different process 23 models in different execution environments. Although process In order to manage the complexity of the process, it was analysis is one step in BPM lifecycle and without process segmented into three sub-sequential sub-processes (Fig. 1.) digitalization it creates small part of the BPM added value for based on a three level SCOR model [18] – order initiation, the business. order production and order completion (Fig. 1). Each of this subprocess was modeled separately by using BPMN 2.0 Application of BPM is popular among large enterprises, modeling language and Camunda Modeler tool [19]. but small-medium sized businesses differ in terms of available resources, process relation, work ethics and the speed of decision making [7]. Thus the application of BPM in SMEs must be investigated. D. Case studies of BPM application in SMEs A case study was conducted in three different SMEs in Fig. 3. Top level order management process model Belgium by C. Bauwens and T. Van Dorpe. They state that the maturity level of an SME must be assessed to understand Order initiation subprocess in target business is not where the organization is with its BPM development [8]. documented or formalized, its specific order of execution is Hammer’s Process and Enterprise Maturity Model [9] and defined by optics sales assistant at the order initiation time for McCormack’s Business Process Orientation Maturity Model each instance separately. Input of this subprocess is client [10] are used to access SMEs under research. Authors needs and output – a filled order. Order initiation subprocess conclude that SMEs have rather low maturity level, they are is explicitly presented in Appendix A, which we think is pointing to weak spots of the small businesses such as lack of enough to show the overall complexity of the underlying documentation and limited inner efforts to process business logic of the analyzed business domain. Order improvements. As the recommendations of improvement production subprocess consists of order manufacturing goals are provided, no details of how to improve process internally or externally. Tasks related to order data sending to models is provided. manufacturers, ordering lenses, sending order to production sites error management and quality control procedures. Order Another case study conducted within Australian Small completion summarizes notifying client, handing finished Business by I. Dallas and M. T. Wynn goes through steps of order to a client, receiving final payment and generating BPM initiative and gives implications on what could be invoices if clients ask. In the scope of order management in improved in the methodology and observations which could optical retail business there were 12 process models created in help other SMEs [2]. Though process evaluation and total. optimization get very little attention as the analyzed business was under establishment. Only some advantages of Business As this process is executed the target business is facing Process Management Systems (BPMS) such as automatic following problems: work allocation are introduced as main benefits for SMEs. The  The states of the orders are not tracked. implementation of BPMS systems in small businesses is also widely discussed in the work of Veldhuizen R., Ravesteijn P.  Order fulfillment date often passes due date. and Versendaal J. They distinguish main differences of SMEs  Late order data retrieval and inaccuracies in it. and large enterprises and suggest an adapted BPMS implementation model [11].  There is no responsible person for each order. As we can see from conducted case studies, despite that  Delays in notifying clients. BPM is a process improvement methodology, there is almost no research in the area of a process evaluation and  Sometimes not all required documentation is filled by the optimization for SMEs. Considering that it is one of the early sales assistants. stages of BPM lifecycle and BPMS creation stages, effective  Long duration of changes implementation. process optimization might save both time and financial resources of such companies.  Close to no control in order management by managers and other business authorities. III. PROCESS OPTIMIZATION IN A SMALL OPTICAL RETAIL BUSINESS B. Process optimization Creating an order management information system based A. Optical retail business order management process on an unoptimized process would be inefficient. Considering The case organization is an optical retail branch business, the size of the optical retail company process optimization which have optical-shops located across Lithuania and is in must be done at minimum cost. For these reasons qualitative the market since 1997. It fits all small business parameters and quantitative analysis will be performed and there will be defined in section 1. Only process of order management of a brief discussion about applicability of each method in SME prescription glasses production will be optimized in the scope under research. As mentioned in [2], the qualitative analysis of this research. Order management is quite complicated in in a small business company might be rejected or seen as this type of business. It always must adapt to dynamic market, redundant. Though without it, it is close to impossible to new technologies and products introduced in the field and conduct process optimization. finally to always changing systems of suppliers. For this reason, rigid, hard-coded or universal off-the-shelf solutions 1) Value-Added Analysis are usually not suitable or are too expensive to deploy and As mention in section 2, the main task of this analysis is support an ongoing process. The current (as-is) order elimination of non-value-adding tasks. The presented order management process does not have a supporting information management process was broken down to a task list of 71 system. basic task. Each of the task was then given an assignee and a 24 category whether it was value-adding, business value-adding source Camunda BPMS platform in order to get highest or non-value-adding task. The analysis was conducted with a amount of added-value from BPM initiative and process supervision of the company CEO. 11 non-value-adding tasks optimization. Two order management sub-process were were identified – most of them related to manual data entry completely digitalized. The main advantages of such system tasks, filling of different forms, work related to a not unified identified by the business stakeholders after the presentation process throughout the company. A plan for each of this task of the solution are as follow: was made – either it was to be automated or eliminated. There were 31 business value-adding tasks which were also revised  Process execution based on an executable process model. and if possible planned to automate by introducing business Process model-based execution ensures that all required process management system. documentation is present during order management and stored after the order is fulfilled. 2) Root Cause Analysis During this analysis only problems stated in presentation  Business rules automation. DMN decision tables are of current order management process were evaluated using integrated in the executable process model and are 6M perspective model presented in section 2. CEO of the supported by the platform. Special offers, discounts for company was involved in all stages of analysis. From the product groups are automatically applied during process obtained results we can see that most of the problems in order execution. Most importantly data defining business rules management were arising from the way the process is be easily changed by an authorized user with no specific executed (method), technical (machine) and human (man) experience in information technologies. Because of this, factors. For readability analysis of each problem was depicted the implemented solution is considered flexible. with cause-effect (Ishikawa) diagrams. Considering each  Tasks allocation and required data presentation. Business problem, process model was revised, and a solution was process management system (Camunda Tasklist suggested. After the analysis process was remodeled, a component) allocates tasks and provides only relevant business process management system and knowledge system task data for the sales assistant at proper time. No were introduced. excessive data is provided, nor it is required to look up Impact assessment and issue register was not created for any data during order management process. because as defined in section 2, in the context of SMEs with  Automation of manual tasks. Most of the manual tasks relatively small amount of problems this method is excessive. such as filling order contract, finding order or client data C. To-be model, simultaion results. and sending notifications to clients were automated and are performed by the BPMS engine. The main reason behind qualitative analysis was to find out how effective qualitative analysis is in terms of process Other advantages of BPMS system such as automatic task execution properties such as execution price, duration and list creation for each employee by task priorities, process resources utilization. Qualitative analysis was conducted by execution data monitoring or the ability to implement changes simulating process models. Process model metrics and to process model with little effort and minimal costs are theoretical models were not applied because of lack of expected to be identified by the business in a long-term testing information and technologies that could be used in SME. As- of the created system. Advantage of information system being is and to-be process models were simulated using opensource built on executable process model is considered an advantage process model simulator BIMP [20]. Two-year orders for the developers or administrators of BPMS as it does not historical data was used as input parameters for this analysis. create direct value for the target business. By making the process unified and removing non-value E. Discussion adding tasks we have made the process 12.9% shorter in duration. The remodeled process included more tasks After conducting qualitative analysis on an order concerning quality control of produced prescription glasses management process of a small optical retail business we have but overall still was 4% shorter in duration and most obtained good results for further process reengineering. From importantly it reduced resource utilization by 15.9%. On the the owners and managers of target business it was expected price point, average execution cost increased by marginal that this type of analysis will not be useful in the scope of 1.32€, this was probably a result of so called devil’s small business. Although it pointed out the redundant tasks quadrangle [21] – by improving process quality and speed, we and weak spots of the process and it was easier to improve the have increased its execution price. The optimization results on process model. For a fluent qualitative analysis there were not process flexibility were not tested. Though considering enough information on how to perform it, though a lot of relatively small rate of process initiations in optical retail definitions of different methods can be found. It is very business flexibility might be more linked to the ability of unlikely that such analysis could be performed in a SME changing process model than adapting to increased amounts business without external consultants. Value-Added analysis of process instances. showed unnecessary tasks, but its results alone lacked information on what parts could be improved. We suggest that D. Process model digitalization and automation by this analysis method should always be used with some introducing Business Process Management System problem-oriented method. In such method, like Root-Cause A process quality analysis alone can have little to no analysis, smaller number of perspectives than in 6M model is impact on the execution of an as-is process in a company. not advised as it may be difficult to point out which parts of Especially, when a process is remodeled with a supporting the process require improvement. Finally, the qualitative information system in mind. In case of analyzed small optical analysis took more effort to complete than expected. This retail business order management process, a business process should be considered and its advantages clarified for the management system is presented which supports to-be process stakeholders as it can be rejected by businesses as not model. Prototype of this system was developed on an open- 25 necessary part of process optimization or automation in an ISBN 978-3-642-33142-8. early stage of initiative. [7] BERNAERT, Maxime, POELS, Geert, SNOECK, Monique and DE BACKER, Manu. Information Systems for Small and Medium-sized Quantitative analysis, as it stands for, gives specific, Enterprises. Information Systems for Small and Medium-sized comparable results that are understandable for all process Enterprises [online]. 2014. No. 2, p. 67–96. DOI 10.1007/978-3-642- stakeholders. For this reason, it is much easier to conduct such 38244-4. Retrieved http://link.springer.com/10.1007/978-3-642- analysis in the scope of SME than qualitative analysis. As 38244-4 expected, the qualitative analysis gave strong backup to the [8] BAUWENS, Cedric and VAN DORPE, Thomas. Business Process results of previously conducted analysis. Although process Management in SMEs. Universiteit Gent, 2018. model simulation requires basic process modeling knowledge. [9] POWER, Brad. Michael Hammer’s Process and Enterprise Maturity Added the usability of the used simulation tool, it is same as Model. Business Process Trends. 2007. No. July, p. 1–4. in the case of qualitative analysis – it is unlikely that this type of analysis can be conducted inside SME business with no [10] MCCORMACK, K. and JOHNSON, W. Business Process Orientation: Gaining the E-Business Competitive Advantage. . 2001. external help. To achieve this a more stable simulation tool must be developed and more information on how to use it must [11] VELDHUIZEN, R., VAN RAVESTEIJN, P. and VERSENDAAL, J. be provided for the user. BPMS implementations in SMEs: Exploring the creation of a situational method. 25th Bled eConference - eDependability: Reliable IV. CONCLUSIONS AND FURTHER RESEARCH and Trustworthy eStructures, eProcesses, eOperations and eServices for the Future, Proceedings. 2012. Vol. 1949, p. 84–98. As we can see from an overview of a business process [12] HEINRICH, Robert and PAECH, Barbara. Defining the Quality of optimization methods and specific optimization case in a Business Processes. Modellierung 2010 P-161 [online]. 2010. small optical retail shop, these methods can be applied in order P. 113—148. Retrieved to optimize small optical retail business order management http://subs.emis.de/LNI/Proceedings/Proceedings161/P- process. Although for these methods to be applicable widely 161.pdf?origin=publicationDetail#page=134 in small businesses there must be broader amount of [13] ROONEY, J J and VAN DEN HEUVEL, L N. Root Cause Analysis information available on this topic. For example, a shared for Beginners. Quality Progress. 2004. No. July, p. 45–53. knowledge base with examples for process optimization [14] EDITORS, Series, BERNUS, Peter and SHAW, Michael J. Handbook should be accessible. Further aim of this initiative is to on Business Process Management 1 [online]. 2010. ISBN 978-3-642- perform a long term BPMS system usability research in a 00415-5. Retrieved http://link.springer.com/10.1007/978-3-642- small optical retail business. Employees attitude and 00416-2 effectiveness of BPMS must be investigated in such [15] DELSANTER, Judith. Six sigma. Managing Service Quality: An environment despite initial advantages recognized by the International Journal. 1992. Vol. 2, no. 4, p. 203–206. stakeholders of the process like process model based DOI 10.1108/09604529210029353. execution, business rules automation, tasks allocation and [16] WILKINSON, Leland. Revising the Pareto Chart. The American automation. Statistician. 2006. Vol. 60, no. 4, p. 332–334. Further research on the topic of process optimization could DOI http://dx.doi.org/10.1198/000313006X152243. be pointed to improving process simulation and modeling [17] VANDERFEESTEN, Irene, REIJERS, Hajo A. and VAN DER tools by making calculations of process model metrics such as AALST, Wil M P. Evaluating workflow process designs using cohesion in real time thus making the evaluation of those cohesion and coupling metrics. Computers in Industry. 2008. Vol. 59, models faster and easier. no. 5, p. 420–437. DOI 10.1016/j.compind.2007.12.007. [18] HUAN, Samuel H., SHEORAN, Sunil K. and WANG, Ge. A review REFERENCES and analysis of supply chain operations reference (SCOR) model. [1] EC. Annual Report of European SMEs 2017-2018. 2018. Supply Chain Management: An International Journal [online]. 2004. ISBN 9789279968228. Vol. 9, no. 1, p. 23–29. DOI 10.1108/13598540410517557. Retrieved http://www.emeraldinsight.com/doi/10.1108/13598540410517557 [2] DALLAS, Ian and WYNN, Moe Thandar. Business Process Management in Small Business: A Case Study. Information Systems [19] CAMUNDA. Modeler. [online]. [Accessed 2019.02.7]. Retrieved for Small and Medium-sized Enterprises [online]. 2014. No. 2, p. 67– https://camunda.com/products/modeler/ 96. DOI 10.1007/978-3-642-38244-4. Retrieved [20] BIMP. BIMP - The Business Process Simulator. [online]. http://link.springer.com/10.1007/978-3-642-38244-4 [Accessed 2018.06.13]. Retrieved http://bimp.cs.ut.ee/ [3] FREUND, Jakob. Camunda BPM Compared to Alternatives [22] REIJERS, H. A. and LIMAN MANSAR, S. Best practices in business [online]. 2015. Retrieved https://network.camunda.org/whitepaper/8 process redesign: An overview and qualitative evaluation of [4] LA ROSA, Marcello. Interview with Michael Rosemann on “The successful redesign heuristics. Omega. 2005. Vol. 33, no. 4, p. 283– Role of Business Process Management in Modern Organizations.” 306. DOI 10.1016/j.omega.2004.04.012. Business & Information Systems Engineering [online]. 2016. Vol. 58, [23] Borowik, G., Woźniak, M., Fornaia, A., Giunta, R., Napoli, C., no. 1, p. 89–91. DOI 10.1007/s12599-015-0419-8. Retrieved Pappalardo, G. and Tramontana, E., 2015. A software architecture http://link.springer.com/10.1007/s12599-015-0419-8 assisting workflow executions on cloud resources. International [5] HARMON, Paul. The Scope and Evolution of Business Process Journal of Electronics and Telecommunications, 61(1), pp.17-23. Management. 2014. ISBN 9783642451003. [6] DUMAS, Marlon, LA ROSA, Marcello, MENDLING, Jan and REIJERS, Hajo A. Fundamentals of Business Process Management. Heidelberg, New York, Dodrecht, London : Springer, 2013. 26 APPENDIX A. ORDER INITIATION SUB-PROCESS The following diagram presents the first sub-process of Order management business process, which is Order initiation. The next two diagrams represent two sub-processes of Order initiation, namely, Choose spectacles frames and Create order. 27