The trap of theoretical regulations and the benefits of Process Mining
Business process automation based on subjective descriptions often fails. Only about 13% of companies are satisfied with their first wave of automation if it was based solely on manager surveys. Attempting to hardcode unoptimized processes only locks in inefficiencies and shadow processes.
To uncover the actual state of affairs, companies leverage Process Mining. It analyzes digital footprints (event logs) in ERP, CRM, and other systems. This allows for precise identification of operational timestamps, loops, and bottlenecks, reducing prioritization errors by up to 49%.
Separating processes and logic: BPMN 2.0.2 and DMN
To ensure architectural flexibility, international standards should be used. BPMN 2.0.2 (ISO/IEC 19510:2013) is used to model and orchestrate process flows. However, a common mistake is attempting to integrate decision-making logic directly into the process diagram.
To address this, the DMN (Decision Model and Notation) standard is applied. Separating business rules into DMN tables allows analysts to update decision logic (such as calculating discounts or limits) independently, without involving developers or altering the overall process structure.
Readiness criteria for process automation
Priority should be given to automating processes that meet the following criteria:
- Stable rules: decision logic does not change erratically and is easily defined via DMN.
- High standardization: the process aligns with the BPMN standard without numerous shadow workarounds.
- Structured data: information is stored in system logs rather than shared verbally or via messengers.
- High ROI potential: high frequency of operations with measurable delays.
ROI calculation and performance evaluation
ROI assessment must be based on hard operational metrics from system logs, such as average step duration and error rates. Comparing the current cost of process execution with the projected post-implementation cost allows for calculating the financial impact before development even begins.
Impact on the industry
For businesses and the IT sector, relying on subjective surveys for automation leads to high failure rates and locked-in operational inefficiencies. Transitioning to data-driven process analysis and standardized architectures is becoming a baseline requirement to avoid wasted development budgets and maintain market agility.
How to prepare
- Analyze actual digital footprints: Implement Process Mining to extract event logs from ERP and CRM systems instead of relying on theoretical regulations.
- Separate process and logic: Use BPMN 2.0.2 for workflow orchestration and DMN for decision rules to ensure the system can be updated without rewriting code.
- Apply strict readiness criteria: Only automate processes that feature stable rules, high standardization, structured data, and clear ROI potential.
Prepared by a Software Ukraine member. Original publication.