A process drawn in a meeting usually shows the intended workflow. Actual records may reveal waiting requisitions, repeated edits and different routes through the same activity. Process mining is an approach for examining that difference through data.
What does process mining examine?
The IEEE Task Force's Process Mining Manifesto describes extracting process knowledge from information-system event records. IEEE Task Force on Process Mining — Process Mining Manifesto The idea is not new. Its relevance to procurement grows as more workflows leave usable digital records across connected systems.
A starting dataset needs a case identifier, an event name and time information. Records such as “created”, “approved” and “transferred to sourcing” can form a sequence for one requisition. A record containing only its current status may not be sufficient to reconstruct earlier waiting periods.
Check data availability first
Identify which system holds each required event. Are time zones consistent? Is the business event time being confused with the data-transfer time? If one requisition is split across several auctions, how is the relationship preserved? A process map built before answering these questions can create a misleading impression of precision.
Start with a small sample. Follow the same transaction through its records and through a conversation with the person who handled it. Classify discrepancies as incorrect data, missing records or genuine process variation. Operational knowledge should accompany the analysis rather than be replaced by it.
Ask questions that lead to action
Go beyond “How many days does it take on average?” Ask which stage accumulates waiting, whether incomplete requisitions travel through more steps and whether technical clarification takes longer in certain categories. These questions support targeted changes instead of blunt measures such as removing every approval.
Do not rely on a single average. A few unusually long cases may distort it. Compare similar categories and scopes, recognising that urgent and routine purchases may follow different rules. Confirm the explanation with the relevant team before changing the workflow.
How Tenflex data may contribute
Tenflex usage and transaction information may contribute to an organisation's analysis within the available export and API scope. Verify access to the necessary event history before planning the analysis. Do not assume Tenflex contains a built-in process-mining module; this article describes a separate analytical exercise.
Choose one question for the first project, such as what delays an approved requisition from reaching sourcing. Identify the required records, validate a small dataset and discuss the finding with the process owner. The value comes from explaining which step should change and why, rather than producing an elaborate picture of the workflow.
