Redesign your processes to run without paper chases

Business Process Automation, in depth

1. P1

   Redesign the process before automating it

   Automating a process exactly as it runs today often preserves its problems. Requests pass through unnecessary approvals, information is entered twice in different systems, and exceptions are handled through emails nobody tracks. Mapping the current process with the people who perform it reveals these issues and the reasons behind them.

   Redesign follows mapping. Ask which steps add value, which approvals reduce real risk and which information is actually used later. Removing a redundant approval level or a duplicate data entry step often saves more time than any technology, and it makes the automated version simpler and cheaper to build.

   Business process automation then covers the whole flow: forms that capture complete information, routing rules, approvals, notifications, document generation and integration with systems of record. This differs from robotic process automation, which mimics clicks in existing screens and suits narrower tasks where systems cannot be changed. Involve the people who do the work throughout. They know where exceptions occur and which shortcuts exist, and their support determines whether the new process is adopted or bypassed with familiar spreadsheets and emails.

2. P2

   Designing approval workflows that work

   Approvals are central to many business processes, from purchases and travel to discounts and contracts. Poorly designed approval chains slow the business down without adding meaningful control. Good design balances speed with appropriate oversight, using the principles below as a starting point.

   Rules should be based on clear criteria, such as amount, department or risk category, so routine requests move quickly while unusual ones receive attention. Delegation during leave and automatic escalation for overdue approvals prevent requests from waiting indefinitely. Mobile approval matters for managers who travel or work away from desks. Approvals through email, Teams, Slack or a mobile app, with enough context to decide without opening other systems, shorten cycle times significantly.

   Platform choice depends on complexity. Low-code platforms suit many departmental workflows, while custom builds suit processes with complex rules, high volumes or deep integration with core systems. Many organizations use both, choosing per process based on complexity and expected lifespan.

   • Approval levels tied to clear thresholds and categories.
   • Delegation rules for absences.
   • Escalation when approvals exceed agreed times.
   • Full context available within the approval request.
   • An audit trail of every decision and comment.
3. P3

   Measuring process performance after automation

   Automation creates data that manual processes never captured. Every request has timestamps for submission, each approval and completion, which reveals where work waits and how long each step really takes. Dashboards built on this data turn process improvement into an ongoing activity rather than a one-time project.

   Useful measures include cycle time from request to completion, time spent waiting at each step, volumes by type and department, rework or rejection rates and service level compliance. Comparing these before and after automation shows the real benefit and builds support for further projects.

   Bottlenecks often move after automation. Once routine steps are fast, the slowest approver or a single overloaded team becomes visible. Process owners can then rebalance work, adjust thresholds or add capacity based on evidence. Smaller, task-level automations often complement full process automation. Connecting apps with workflow automation tools handles notifications and data syncs around the main process, while the core workflow keeps control, approvals and audit trails.