How Digital Automation Ensures Efficient Processes.
These days, companies are constantly challenged to adapt quickly and efficiently to ever-changing market conditions. Powered by a wide range of modern technologies and an increasing level of networking, digital transformation is of great importance in this context. The term, however, stands not only for digitization but also for another important area of digital transformation, namely automation.
The automation of a business process is arguably always also its digitization. On the other hand, not every digitized process is also an automated process. So, what constitutes automation? And why is it so important?
Identifying a process
Existing processes should be monitored on an ongoing basis and checked for their cost-effectiveness. Does the current process still provide the expected added value? Or should it be adapted, revised or replaced altogether? Possible specific questions to ask about existing processes are:
- Are there media discontinuities that can be eliminated by integrating the surrounding systems?
- Are there performance indicators for throughput times, frequencies, or error rates?
- Are there any analyses of customer satisfaction and customer complaints?
- Are there changes in the market situation making it necessary to modify the process?
- What is the process' priority compared to other processes?
- Which factors can be improved by means of IT solutions or IT support?
A typical case involves a company that has always had an undocumented, cumbersome accounts payable process. This process is executed multiple times a day and takes a lot of time. It is also error-prone and processing time is dependent on employees. These weaknesses are exacerbated by frequently repeated process runs.
Understanding the existing process
Should the existing process be analyzed first or should we switch to the target process right away? Both approaches are possible, whereby a distinction should be made between process optimization and process re-engineering. In re-engineering, the existing process is ignored and consideration is given to how the optimized process would have to run, applying current technologies. In continuous process optimization, one looks at the existing process and uses techniques from Lean Six Sigma, for example, to gradually achieve improvements. Since, in our case, we want to incorporate the findings from the existing process into the new process, we will first focus on the existing process and try to find available sources of information along with usable data. Possible sources are: The process map to locate the process in the process landscape, process models and process documentation, a survey of process owners and process participants, and a data analysis through process mining.
To get a picture of the invoicing process we designate individual work steps and depict them in their chronological sequence in the form of an end-to-end process model. For this purpose we choose the BPM Notation 2.0 which is established as today's standard.
The model shows the process flow in an overview, i.e. it shows a start and an end event, with a maximum of ten work steps, without branches and other secondary processes. Work steps are assigned to the responsible process participants.
Let the process analysis reveal weaknesses
The next steps should be taken in close cooperation with the parties involved in the process. The aim is to uncover any weaknesses in the process. To this end, metrics, quantity structures, and other data relating to the process should be evaluated. It will also be useful to specifically check the process with the questionnaire previously used in process identification. This will reveal the following deficiencies in the existing accounts payable process:
Long throughput times, as (paper) invoices are often delayed at the approval points; many media discontinuities, for example in invoice verification, posting, and archiving; a high error rate, as process steps are performed manually; and low scalability, as each process run means the same amount of work.
This already demonstrates that a significant improvement in efficiency and reliability can ultimately only be achieved with IT support. To this end, it is necessary to be able to receive and process invoices along with their associated user data in digitized form. This is a first step towards digitization of the process. However, what we are aiming for as an end result of our process improvement is an IT system that can independently start the process, control it, and also monitor it.
The target process at a glance
Using a SIPOC analysis ( supplier, input, process, output, customer), start and end points of the target process are determined first and process steps in between are recorded in a process diagram. This should achieve the following: Demarcation from other processes, definition of input and output parameters of the process, a uniform understanding of the basic flow of the process, as well as the responsibilities and the information processed and transferred in the process. The chart represents only the principal flow of the process, without branches and error handling. The activities themselves merely describe the summary of one or more process steps of the responsible process participant.
The target process is triggered by every received invoice, which must already be available in scanned form. This clearly sets us apart from a separate, upstream process that handles processing and digitization of incoming invoices. Payment processing is also out of scope, as it involves other actors and systems being part of a downstream process. Furthermore, it becomes apparent that the transferred information also includes the vendor master data, usually managed and made available in one of the surrounding systems.
The target process in detail
Now the task is to detail individual activities for the target process. The aim is to answer the following questions: Who is involved in the process and what activities do they perform? What interactions take place between process participants? What interactions take place between process participants and surrounding systems? Are there any exception and error handling procedures?
Usually, there exist several business cases within one process, each of which is considered, understood, and recorded separately from the beginning to the end of the overall process. Ultimately, this results in a process diagram that contains not only the main process flow, but also all alternative process flows and their branches. Processed and shared data and information are also recorded.
During this phase of the process survey we do not initially distinguish which of the processes are to run manually, with IT support, or fully automated, since we first want to understand and optimize the process from a purely organizational perspective. The focus is thus entirely on those involved in the process who should understand and help shape their own workflows and also follow them later.
First, the accounting department reviews the newly received invoice, completes it where necessary, adds account assignments and also determines the releasing department. If the invoice contains errors, the accounting department can reject it, which means that the invoice is archived and the process ends prematurely. As the process continues the invoice with the enhanced data is forwarded for review and approval. If the responsible parties reject it the invoice is again assigned to accounting for post-processing. Once the approvals have been obtained the invoice is checked one last time by the accounting department, the transaction is executed, and the invoice is archived with the associated approval log. This completes our accounts payable workflow and the invoice is ready for further payment processing.
The target process digitized and automated
As we could see in the process analysis the target process should not only be accurate but also efficient and reliable. Therefore, the process which is now presented in detail from an organizational point of view is to be optimized by means of additional IT support. The resulting target process should ultimately be automated as far as possible, i.e., it should be able to be processed purely by machine.
For this purpose we introduce a central software component that is responsible for the automated control of our target process: the workflow engine (WE). This is made possible by converting our existing process diagram into a technically executable BPMN diagram that can be directly interpreted and executed by the WE. Thus, the WE has complete control over the process flow and performs the following tasks: Starting new process instances with each incoming start event, controlling all modeled flows and branches, assigning pending tasks to process participants and notifying them, executing service calls and script tasks, and managing process data.
The transfer from the detailed target process model to the technically executable process model is done systematically by summarizing the individual work steps of each process participant and representing them as a human task. This will ultimately result in a BPMN diagram representing the interaction of machine-controlled processes with human work steps.
As can be seen from the technical model, we have introduced an additional system lane for all service calls. These are tasks that can be executed entirely without human interaction. Through the WE, we also have the ability to introduce an escalation path at the approval points, which is executed whenever an invoice fails to be approved within the specified time period.
Process improvements through a systematic approach
Process automation does not imply that the entire process will run fully automatically. Rather, the entire process is managed by the workflow engine, i.e. started, controlled and also ended again. Due to the technically executable process model the WE knows at any time which path it has to follow and which task is currently pending - a human work step or a service call (for example to a peripheral system). By giving the workflow engine full control over every running process, we can achieve several significant improvements at once:
Effectiveness: As soon as a certain start event occurs, a new process instance will be automatically started and the next pending task will be assigned to the responsible persons (or system). Completed tasks are evaluated immediately and the process is executed accordingly. This significantly reduces process cycle times.
Efficiency: Automation also provides the option of parallel processing individual work steps or even entire sub-processes in the process flow.
Reliability: The process is executed and monitored according to predefined rules. Human work steps can be limited and media discontinuities can be eliminated. This reduces sources of error even further.
Scalability: The workflow engine is designed to control and handle many process instances running simultaneously. This means that even an increased workload can be handled with the same resources.
Error handling: Errors occurring during process handling can be intercepted and addressed in a targeted manner. Conditional process paths can be used to implement exception or escalation scenarios.
Integration: Tasks can be outsourced or process data exchanged via service calls to surrounding systems.
Agility: The process can be modified and adapted to new requirements at any time. Whether it is a changed process flow, the addition of process participants, or the integration of further peripheral systems: the workflow engine is not affected by this and will remain unchanged as a software component.
Monitoring: The workflow engine can permanently monitor every running process instance and ensure reliable operation. Process history and key figures can also be evaluated at any time in order to further adapt and optimize the existing process.
Are you on your way to digital transformation?
We would be happy to accompany you as an experienced and reliable partner, from the development of your digital strategy and the survey of automation potentials to the modeling and automation of your target processes and the operation and continuous improvement of your business processes.