Structured process optimization in banks and insurance companies with/without BPM

In several digitization projects, especially in the consulting environment, we initially started with classic BPMN engines such as Camunda, Flowable, appway, Axon Ivy, … based on the assumption that the underlying business process follows a clearly defined, directed flow. The aim is often to achieve structure, transparency and automation through the use of a BPM engine.

In practice, however, it has repeatedly been shown that many processes, especially in personal advice, are not strictly sequential. Instead, users require a high degree of flexibility in navigation, e.g. to jump between individual modules or subject areas (investment advice, pensions, financing, ….), to process certain sections several times or to respond to customer needs depending on the situation.

Although approaches such as case management have been discussed or implemented in some cases, they were often unable to satisfactorily reflect the desired openness and user-friendliness. As a result, we deliberately decided against a BPMN-based implementation in several projects and instead opted for lightweight solutions such as single-page applications with simple state logic. The few necessary rules and process dependencies were each implemented with minimal, self-developed state machines.

Our insight: not every business process benefits from a BPM engine. Especially for non-linear, user-driven processes, individual, simple architectures with a clear API structure and modular design can be the better choice – both in terms of user experience and development effort.

Conclusion: From BPMN to modularity: architectures for dynamic interactions

Our experience clearly shows that when processes are dynamic, user-driven and context-dependent, classic BPMN approaches quickly reach their limits. Instead of investing in complex modeling and orchestration logic, it can make more sense to rely on simple, modular architectures that offer maximum flexibility. The targeted use of APIs, reusable components and minimal state machines results in solutions that are not only technically maintainable, but also remain intuitive and customizable for users. The decision not to use a process engine is not a step backwards, but often a step towards greater agility, clarity and future viability.

But: BPM is justified, especially when processes need to be clearly defined, comprehensible and rule-based. We have been able to utilize the strengths of BPMN tools very successfully in many other project contexts. A few examples:

Revision management in a document information system

• Record, check and approve changes to regulatory documents

• Merge and release revisions

• Automated notifications about new versions
  → BPM offers clear processes, traceability and high governance here

Digital customer onboarding

• From entering customer data to risk assessment and opening an account

• Many decisions, mandatory steps and dependencies
  → A prime example of BPM, as structured processes are necessary

Infrastructure monitoring with automated troubleshooting

• Monitoring of IT systems

• Automatic triggering of escalations and replacement orders

• Planning and coordination of interventions
  → BPM reliably controls the process and ensures a reliable response

Internal software lifecycle in a Swiss bank

• Capture, bundle and form requirements into releases

• Coordinate role-specific approvals, testing and deployment
  → Complex processes with many participants – BPM creates transparency and efficiency

Approval processes with dual control principles

• Release of payments, contracts or critical changes

• Adherence to regulatory requirements and compliance requirements

• Automated escalation in the event of missing approval or missed deadline
  → BPMN is ideal for traceable and role-based approval processes

These examples show: BPMN is always worthwhile when processes need to be structured, repeatable and auditable. The key is to choose the right tool for the respective context with a sense of proportion and technical understanding.