Video summary

FBPM-5.2.: Fundamentals of Business Process Management (BPM) - Process Discovery Methods

Main summary

Key takeaways

Educational

Main ideas: Process Discovery Methods in BPM

The video explains several complementary ways to understand and model a business process (i.e., process discovery). For each method, it discusses what it does well, common pitfalls, and how to combine methods in a practical sequence. It also notes that workshop-based discovery may vary depending on organizational context.

Methods of process discovery (with key details)

1) Document analysis

Purpose: Learn about a process by reading existing written materials.

What to look for in documents:

  • Documentation that describes:
    • Roles involved
    • Activities to be conducted
    • Business objects used
  • Related documents that aren’t purely process descriptions but are still informative, such as:
    • Internal policies
    • Organizational charts
    • Employment plans
    • Quality certificate reports
    • Glossaries and handbooks
    • Forms
    • Work instructions

How it’s used:

  • The process analyst reads documents before interacting with domain experts.
  • This helps the analyst familiarize themselves with process details and ask more effective, targeted questions—making expert interaction more efficient.

Common challenges:

  • Documents may be not process-oriented (requiring abstraction/refinement).
  • Documents may be inaccurate (not fully aligned with the real process).
  • Documents are often outdated, reflecting how things worked in the past rather than current reality.

2) Observation

Purpose: Understand how the process actually runs by watching people perform it.

Two observation roles are described:

  • Active role observation

    • The analyst plays a specific role in the process (e.g., acting as a customer).
    • Benefit: firsthand insight into what happens.
    • Limitation: may not reveal the whole span of the process end-to-end.

  • Passive role observation

    • The analyst observes others handling work in the process.
    • Best practice: trace the business objects that flow through the process.
    • Limitation: observation can change participant behavior (people may work “more accurately and faster”).
    • Result: understanding may be limited and less representative of normal operation.

3) Automated process discovery (process mining)

Purpose: Infer a process model automatically from system event data.

Core concept / inputs:

  • Uses event streams and data stored in information systems
  • Organizes data into an event log
    • The event log captures how different cases are processed

Output:

  • Automatically discover a BPMN process model
    • The model reflects what is visible to the information systems (i.e., operational reality as captured by data)

Additional analysis techniques mentioned:

  • Conformance checking (if you already have a process model)
    • Identifies deviations between the model and actual behavior
  • Enhancing an existing model with frequency information
    • How often certain behavior occurs

4) Interviews

Purpose: Capture not only how the process works, but especially why.

Why interviews matter:

  • Domain experts can explain:
    • How things work
    • Why they work that way (described as extremely important)

How interviews are used:

  • Adjust process models based on insights
  • Validate understanding by confirming the process is understood correctly

Interview structuring options:

  • “Start to end” (from start of process toward the end)
  • “Results to causes / backwards step-by-step”
    • Start from what the process produces, then ask how outcomes are achieved

Important requirement: shared terminology

  • Analysts must ensure terminology matches what domain experts mean.
  • Abstract terms (e.g., “process,” “project”) may be understood differently.

Focus on exceptions:

  • Normal behavior is often discussed; exceptional behavior may be neglected.
  • Analysts should explicitly ask:
    • What happens when something goes wrong?

5) Workshops

Purpose: Reassemble shared understanding collaboratively with stakeholders.

How workshops work:

  • Gather key stakeholders/participants (domain experts; potentially a process owner; tools may be used).
  • The process analyst acts as a facilitator.
  • Participants build the process together using:
    • Brown paper / physical artifacts, or
    • A modeling software tool

Preparation and effort required:

  • Workshops require significant time and effort to prepare and conduct.
  • A typical format described:
    • 3 to 5 half-day sessions to reach suitable process detail
  • Additional overhead:
    • Scheduling so all stakeholders can attend.

Challenges of process discovery (context effects)

The video contrasts two hypothetical settings and how they may influence workshop effectiveness:

Company A

  • Young, rapidly grown; >100 employees; not yet systematically organized
  • Workshop advantages:
    • Participants may be used to open discussion and collaborative problem-solving.
  • Workshop challenges:
    • Less structured organization can make the task more fuzzy.

Company B

  • Government-owned; extensive health & security regulations
  • Workshop advantages:
    • Likely more structured processes/cultures.
  • Workshop challenges:
    • May lack a strongly discussion-oriented culture, so participants may be less comfortable or less inclined to debate/brainstorm.

Suggested overall sequence / approach (method combination)

A recommended starting and progression is described:

  1. Start with document analysis
    • Provides structured information before stakeholder interaction.
  2. Then observe and use automatic discovery techniques
    • Observation requires less effort from domain experts (who are scarce).
    • Automatic discovery also reduces the burden on domain experts.
  3. Then conduct interviews
    • Interviews provide critical “why” and help validate understanding.
    • Must be neatly prepared because domain experts have limited time.
  4. Workshops as a possible add-on
    • Can be very effective, but may be difficult to organize due to the challenge of bringing all domain experts together.

Speakers / Sources featured

  • No specific individual speakers are identified in the provided subtitles (no named speakers or on-screen sources mentioned).

Original video