Business Process Management Main Summary

• Core processes

Directly tied to value creation

Support processes

Enable core processes

Management processes

Oversee and coordinate other processes

Reference models

models (e.g., SCOR for supply chains, SAP reference architecture) provide reusable templates and best practices for process identification

Process Portfolio

Visual tools (like heat maps or matrices) help in selecting which processes to improve first. The goal is to create early success stories and learn from them.

Automated Process Discovery

• Goal: Automatically generate a process model from an event log.

• Input: An event log.

• Output: A process model that reflects actual behavior.

• Usage: Useful during early discovery or continuous performance monitoring.

• Methods are discussed in detail in Chapter 11.4.

Conformance Checking

• Goal: Compare actual process executions (in event logs) with a given model or set of business rules.

• Input: Event log + process model (or rules).

• Output: Deviations between expected and actual executions.

• Example: If task B is supposed to follow task A but doesn't in some cases, this may signal a problem or an unmodeled exception.

Performance Mining

• Goal: Overlay performance information on a process model to analyze issues such as bottlenecks or delays.

• Input: Event log + process model.

• Output: Enhanced model with performance metrics (e.g., color-coded delays).

• Use case: Identify why a process is slow or where time is being wasted.

Variants Analysis

• Goal: Compare two event logs representing different process outcomes (e.g., successful vs. failed cases).

• Input: Two event logs.

• Output: Differences in process paths or behaviors.

• Use case: Diagnose why certain cases end with complaints or delays versus satisfactory completion.

Event Log

• A structured record of events, typically stored as a table or in the XES (eXtensible Event Stream) format.

• Each event contains:

1. Case ID: Which instance (e.g., which order or claim) the event belongs to.

2. Activity Name: What was done.

3. Timestamp: When the activity occurred.

4. Optional Attributes: Resource (who did it), cost, status, domain-specific data, etc.

Why process mining is important

• Process mining bridges the gap between high-level monitoring and detailed operational insight.

• It enables fact-based decision-making, reduces guesswork, and supports continuous improvement.

• It feeds into both exploratory and question-driven analysis approaches in real-world BPM practice.

Qualitative process analysis

focuses on uncovering inefficiencies, waste, and issues in business processes from a non-numeric,

Value-Adding (VA)

Directly benefits the customer (e.g., repairing a machine).

Business-Value-Adding (BVA)

Doesn’t benefit the customer directly but is needed for compliance or risk management.

Non-Value-Adding (NVA):

Pure waste (e.g., unnecessary approvals or manual data entry).

Waste Analysis

Identifies seven types of waste in processes, organized into categories:

Stakeholder Mapping

Identifying who interacts with the process and their roles.

Issue Register

A living document of observed process issues

o Pareto Analysis

Uses the 80/20 rule to focus on the most impactful issues.

PICK Charts

: Classifies issues based on ease and payoff (Possible, Implement, Challenge, Kill).

o Cause-Effect Diagrams

(Fishbone/Ishikawa): Categorizes causes (e.g., Man, Machine, Method, etc.).

o Why-Why Diagrams

Asks “why” repeatedly to trace a problem back to its root.

o Causal Factor Charts

o Causal Factor Charts

o Cycle Time (CT):

Avg. time from start to end.

o Processing Time

Time actively working on a task.

o Processing Time:

Time actively working on a task.

o Waiting Time

Idle or queue time

Queueing Theory

• Models how tasks wait for resources.

o Arrival rate

o Service rate

o Number of servers/resources

• Simulation

: Empirical modeling to reflect real or imagined process behavior.

Operational Dashboards

• Target Audience: Process participants, operational managers, and process owners.

• Purpose: Support short-term decision-making and real-time process monitoring.

• Focus: Ongoing or recently completed cases.

• Common Metrics:

o Number of active cases (Work-in-Process)

o Case status breakdown: on-time, overdue, or at risk

o Pending tasks by resource (e.g., how many tasks assigned per worker)

• Example: A dashboard in a warehouse displays the number of deliveries due soon using pie charts and histograms, helping dispatchers prioritize urgent orders.

Tactical Dashboards

• Target Audience: Process owners, functional managers, business analysts.

• Purpose: Provide performance insights over a longer period (weeks to quarters).

• Use:

o Detect bottlenecks

o Analyze variability

o Investigate root causes of process inefficiencies

• Metrics:

o Cycle time, waiting/processing times

o Cost per case, resource utilization, defect rates

o Visuals: histograms, drill-downs per task, cross-sectional (e.g., by region), or longitudinal views (e.g., month-over-month trends)

• Use Case: Comparing lead-to-order process performance across different sales regions to identify top-performing teams.

Strategic Dashboards

• Target Audience: Executive managers and senior leadership.

• Purpose: Support strategic decision-making by aggregating KPIs across processes and business units.

• Aggregation Levels:

o Across processes in a process architecture

o Across multiple metrics into a single performance score per dimension (e.g., efficiency)

• Example: An insurance firm uses weighted averages of efficiency metrics from claims processes to generate an overall dashboard view using a balanced scorecard approach.

Tools for Dashboard Creation

• Many BPMSs provide dashboards by default (e.g., Bizagi, Perceptive).

• Standalone tools like Power BI, QlikView, and Tableau allow custom dashboard creation.

• Emerging Trend: Integrating machine learning to predict performance and automatically detect issues (e.g., Nirdizati

the Devil’s Quadrangle

Time, Cost, Quality, and Flexibility.

o Dotted charts

: Show when events happen (useful for spotting clusters or performance bands

o Timeline charts

Show duration of each task, split into waiting and processing time (help detect bottlenecks).