Introduction

Many organisations have detailed event logs that record how work moves through a process. These logs appear in systems such as CRM platforms, IT service management tools, ticketing systems, ERP workflows, and project trackers. Most teams use logs to measure volumes and turnaround time, but logs can reveal something equally important: who collaborates with whom, where work gets handed off, and which resources act as connectors between teams. Social network discovery is a method from process mining that turns event logs into a network map of collaboration and hand-offs. For professionals building structured problem-solving skills through a business analyst course, this technique offers a practical way to move from assumptions about teamwork to evidence-based insights.

What Social Network Discovery Means in Process Mining

Social network discovery uses event log data to infer relationships among resources-people, roles, teams, or systems-based on how they interact within process instances (cases). A “case” could be a customer complaint, a purchase order, a loan application, or an incident ticket. Each case produces a sequence of events, and each event usually includes at least:

• Case ID (unique identifier of the work item)
• Activity name (what happened)
• Timestamp (when it happened)
• Resource (who or what performed it)

By analysing patterns across many cases, you can create a network graph where nodes represent resources and edges represent relationships such as hand-offs, shared work, or joint involvement. The result is a data-driven view of operational collaboration.

Key Relationship Types You Can Extract from Event Logs

Different definitions of “working together” produce different networks. The most common ones are:

1) Handover-of-Work Network

This network captures direct hand-offs. If Resource A completes an activity and the next event in the same case is performed by Resource B, that counts as a handover from A to B. Aggregating this across cases shows frequent handover paths.

This is especially useful for identifying bottlenecks created by excessive routing, unclear ownership, or repeated escalations.

2) Working-Together (Co-Participation) Network

Here, two resources are considered connected if both worked on the same case, regardless of sequence. The stronger the overlap across cases, the stronger the connection.

This helps reveal informal collaboration patterns, such as which teams frequently co-handle complex cases.

3) Subcontracting or Delegation Patterns

In some processes, one resource initiates or “owns” a case but repeatedly delegates tasks to another. This can appear as a repeated pattern of A → B hand-offs.

It can signal healthy role specialisation-or hidden dependency risk if too much work relies on a few individuals.

4) Reassignment and Escalation Signals

If the same activity is performed by different resources in the same case (e.g., reassigned approvals), you can infer instability in task ownership.

High reassignment rates can indicate training gaps, unclear rules, or workload imbalance.

How to Build a Social Network from Event Logs

A clear method matters because small choices change the network drastically.

Step 1: Validate Log Quality and Define “Resource”

Decide what a node represents: individual employees, role names, teams, or automated systems. If names are inconsistent (e.g., “John S.” vs “John Smith”), standardise them first. Also check whether system accounts should be excluded.

Step 2: Choose the Relationship Rule

Pick the network type based on your operational question:

• Investigating delays between teams → handover-of-work
• Understanding collaboration clusters → working-together
• Tracking dependency and delegation → subcontracting patterns

Be explicit and document the rule so stakeholders trust the output.

Step 3: Add Edge Weights and Thresholds

Edges typically carry weights such as:

• Count of handovers
• Percentage of cases involving both resources
• Average delay between hand-off events

To keep the network readable, apply thresholds (e.g., only show relationships appearing in at least 30 cases). Without thresholds, networks can become visually noisy and analytically less meaningful.

Step 4: Interpret with Network Measures

Basic network metrics can convert a graph into actionable insights:

• Degree centrality: who is connected to many others (high interaction load)
• Betweenness centrality: who sits between groups (key coordinators, potential single points of failure)
• Communities/clusters: natural team-like groupings that may or may not match the formal org chart

This is where a business analysis course becomes valuable, because the output needs interpretation in business terms, not just technical visuals.

Practical Use Cases and Decisions You Can Support

Social network discovery is not just an academic exercise. It supports real operational improvements:

• Reducing hand-off overhead: Too many handovers can inflate cycle time. The network shows where unnecessary routing occurs.
• Spotting dependency risk: If a few individuals have extremely high betweenness, their absence could disrupt delivery.
• Improving role clarity: Reassignments and repeated escalations may point to unclear ownership rules.
• Supporting staffing decisions: You can identify overloaded collaboration hubs and redistribute work more evenly.
• Validating process redesigns: Before-and-after networks show whether a change reduced cross-team churn.

Common Pitfalls to Avoid

Like any data-driven technique, social network discovery can mislead if done casually:

• Assuming sequence implies collaboration: A handover is not always cooperation; sometimes it is forced routing. Combine findings with interviews.
• Ignoring time gaps: A handover after 2 minutes differs from a handover after 2 days. Include delay metrics where possible.
• Overlooking automation: System accounts can dominate networks and hide human collaboration unless filtered carefully.
• Privacy and ethics risks: Networks can look like performance rankings. Use aggregated or role-level views when appropriate and communicate intent clearly.

Conclusion

Social network discovery turns event logs into a practical map of how work actually flows between people and teams. By extracting hand-offs and collaboration patterns, organisations can identify unnecessary routing, dependency risks, and opportunities to streamline coordination. The technique is especially useful when paired with business context, stakeholder input, and careful data preparation. For learners and practitioners coming from a business analyst course or a business analysis course, this approach offers a structured way to convert operational data into insights that improve process efficiency and team effectiveness.

Business Name: ExcelR- Data Science, Data Analytics, Business Analyst Course Training Mumbai
Address: Unit no. 302, 03rd Floor, Ashok Premises, Old Nagardas Rd, Nicolas Wadi Rd, Mogra Village, Gundavali Gaothan, Andheri E, Mumbai, Maharashtra 400069, Phone: 09108238354, Email: enquiry@excelr.com.