Process Mining and Lean Six Sigma: The Next Frontier in Process Optimization

For decades, businesses have relied on Lean Six Sigma (LSS) to refine their processes, eliminate waste, and improve efficiency. By combining Lean’s focus on streamlining workflows with Six Sigma’s statistical approach to reducing variation, organizations have made measurable improvements in quality and cost reduction. However, traditional Six Sigma methodologies often struggle with a key limitation: reliance on manually collected data and subjective process mapping. The digital age demands something faster, more precise, and rooted in real-time insights.

This is where Process Mining enters the equation. By extracting and analyzing process execution data from enterprise systems such as SAP, Salesforce, and many more, Process Mining offers an objective view of how work flows through an organization. When combined with Six Sigma’s rigorous approach to measuring process capability, businesses can move beyond static snapshots of performance and embrace continuous, data-driven process optimization.

The Role of Process Capability in Process Optimization

A fundamental aspect of Six Sigma is assessing whether a process consistently produces outputs within predefined performance thresholds. This is where process capability indices – such as Cp and Cpk – become critical. These statistical measures determine whether a process is both stable and capable of meeting customer or business specifications. A Cp value above 1.33, for instance, generally indicates a well-controlled process, while a Cpk value below 1 suggests that outputs are drifting away from their intended targets.

In theory, Six Sigma practitioners use these indices to detect inefficiencies and drive improvements. In practice, however, they often rely on limited, sample-based measurements, which can mask underlying variability and fail to capture the full complexity of a business process. Traditional process capability studies might examine a subset of invoices, manufacturing runs, or service tickets, but they rarely account for the full end-to-end process, across all variations, in real time.

Process Mining changes this paradigm. Rather than analyzing a handful of sample cases, it pulls from complete datasets – every transaction, approval, delay, and rework instance recorded in an organization’s digital footprint. This allows for a true, system-wide process capability analysis, revealing inefficiencies that would otherwise remain hidden. Instead of waiting for quarterly reports on process deviations, businesses can track Cp and Cpk scores dynamically, identifying emerging risks as they happen.

Beyond Static Dashboards: A New Era of Process Analytics

One of the most immediate benefits of integrating Process Mining with Six Sigma is the transformation of Six Sigma dashboards. Traditionally, these dashboards present aggregate statistics – defect rates, cycle times, efficiency metrics – without offering clear visibility into the root causes of process variation. They tell executives that something is wrong but not necessarily where or why.

By embedding Process Mining into these dashboards, organizations can move from descriptive analytics to diagnostic and predictive analytics. Instead of merely showing that a process has drifted out of control, dashboards can now visualize the exact points of deviation, highlight rework loops, pinpoint approval bottlenecks, and quantify the impact of process variations on Cpk values.

Take, for example, a global supply chain operation struggling with delays in order fulfillment. A traditional Six Sigma report might indicate that order lead times exceed the target 85% of the time, with a declining process capability score. However, such reports typically leave analysts guessing whether the issue stems from supplier delays, internal processing inefficiencies, or system misconfigurations.

Process Mining changes this by reconstructing the actual order fulfillment workflow from raw system logs, allowing analysts to see where delays occur most frequently. It might reveal that unwanted additional events or skipped activities (e.g. Check Batches - +20 days) lead to a high negative lead time impact. Instead of generic recommendations for process improvement, businesses can make precise, data-driven interventions.

Revolutionizing the DMAIC Cycle with Process Mining

For decades, businesses have relied on the structured DMAIC methodology – Define, Measure, Analyze, Improve, and Control – to eliminate inefficiencies and enhance operational performance. While the methodology has yielded significant improvements, its effectiveness has often been constrained by the reliance on manual data collection, subjective assessments, and infrequent process reviews. Enter Process Mining, a technology that transforms the DMAIC framework from a methodical, stepwise exercise into a dynamic, data-driven approach that enables continuous improvement at an unprecedented scale.

At the outset of a Lean Six Sigma project, the Define phase is crucial for setting the foundation of the improvement effort. Traditionally, teams spend weeks mapping out business processes through stakeholder interviews and workshops – exercises that often introduce bias and inaccuracy. Employees describe how processes should work, rather than how they actually function. Process Mining removes this layer of assumption by reconstructing the real flow of activities from system event logs. Instead of speculating about inefficiencies, organizations can see precisely where delays, deviations, and unnecessary rework occur. This data-driven clarity not only accelerates project timelines but also ensures that improvement efforts target real bottlenecks rather than perceived ones.

A more significant transformation occurs in the Measure phase, where businesses attempt to quantify process performance. Historically, Lean Six Sigma practitioners have relied on small sample sizes and periodic reporting to calculate key performance indicators such as cycle times, defect rates, and process capability (Cpk). These methods provide only a static snapshot of operational efficiency, missing daily variations and underlying trends. Process Mining eliminates this limitation by tracking every process instance in real time, offering a continuous measurement system rather than an intermittent one. Organizations can observe how process performance shifts throughout the day, the week, or even the year, enabling them to spot patterns that would be impossible to detect with traditional sampling methods.

The next phase, Analyze, is where the contrast between conventional Six Sigma practices and modern Process Mining techniques becomes even more apparent. Lean Six Sigma practitioners have long relied on tools such as Fishbone Diagrams, Pareto Charts, and Regression Analysis to identify root causes of inefficiencies. While effective, these methods are heavily dependent on the expertise of analysts and can be labor-intensive. Process Mining automates much of this investigative work. Advanced algorithms sift through millions of process instances, highlighting variations correlated with delays and defects. The technology can reveal, for instance, that orders processed on Fridays take 20% longer than those on other weekdays, pointing to staffing constraints or system slowdowns. This kind of insight enables businesses to move beyond educated guesses and make process improvements backed by irrefutable evidence.

The Improve phase is where businesses traditionally implement changes, often through pilot programs or manual process redesigns. Yet even the best-designed solutions carry an element of uncertainty – organizations must wait weeks or months to assess whether an improvement delivers results. Process Mining introduces a radically different approach by enabling businesses to simulate process changes before implementation. Instead of making blind adjustments, organizations can test different scenarios in a digital twin of their process. If an analysis reveals that reducing approval steps could cut cycle times by 25%, teams can model the change and assess its impact before disrupting day-to-day operations. This predictive capability drastically reduces the risk of ineffective process modifications, ensuring that only the most impactful changes are implemented.

However, the most profound shift comes in the Control phase. One of the most common challenges in process improvement is sustaining the gains over time. Many Lean Six Sigma projects achieve short-term successes, only for performance to deteriorate once the focus shifts elsewhere. Traditional control mechanisms, such as periodic audits and compliance reviews, often fail to detect early signs of process drift. Process Mining changes this dynamic by introducing real-time process monitoring. Instead of waiting for quarterly reports, businesses can continuously track process KPIs, flagging anomalies as they occur. If a newly optimized process begins to slip – if cycle times creep upward or defect rates increase – automated alerts trigger immediate investigation, allowing organizations to intervene before small inefficiencies snowball into major setbacks.

By integrating Process Mining with DMAIC, businesses are not simply enhancing a well-established framework; they are fundamentally redefining how process improvement works. No longer constrained by limited datasets and retrospective analysis, organizations can operate in a world where process performance is measured in real time, inefficiencies are uncovered automatically, and improvements are continuously validated. This fusion of Six Sigma’s structured methodology with Process Mining’s analytical power represents the future of operational excellence – one where continuous improvement is not an aspiration but a data-driven reality.

Case Example: Optimizing Claims Processing

Let’s consider a global insurance company that struggles with inconsistent claims processing times – a common issue in the financial services sector. The company found that claims were frequently delayed beyond the specified target of 48 hours, resulting in poor customer satisfaction and regulatory penalties. Traditional Six Sigma tools revealed that the company’s claims process had a declining Cpk score, but the root causes of delays were unclear.

By integrating Process Mining, the company was able to gain deeper insights. Process Mining pulled data from the company’s claims management software and reconstructed the end-to-end process, from initial claim submission to final approval. The insights were revealing:

• A significant number of claims were delayed due to missing documentation or incomplete forms, which triggered a lengthy rework cycle.
• Claims handled by certain regional offices had longer cycle times due to a bottleneck in the manual review process. In one specific office, claims were being reassigned multiple times before reaching a final decision, leading to wasted time and inefficiencies.
• Some claims, particularly those involving high-value or complex cases, were subject to excessive approvals, slowing the process further.

With this comprehensive visibility, the company was able to pinpoint exactly where delays were occurring and why. The insurance company could now directly link process deviations to poor Cpk scores, indicating that these inefficiencies were causing the process to fall out of specification.

The next step was to redesign the claims handling workflow. For example:

• The company streamlined its documentation verification step to reduce rework.
• They automated routine approvals, reducing the number of manual interventions in the process.
• The regional offices identified as outliers were provided with additional training and technology support to standardize their procedures.

As a result of these data-driven interventions, the company achieved:

• A 25% reduction in claim cycle time.
• A Cpk improvement from 0.9 to 1.5, indicating a much more stable and predictable process.
• A 30% reduction in customer complaints related to claim delays.

This case demonstrates how Process Mining can elevate the power of Lean Six Sigma by providing granular, real-time data that reveals not just that there’s a problem, but also the precise factors contributing to it. Instead of relying on periodic Six Sigma reviews, the company could now continuously monitor the claims process and intervene as soon as performance deviated from desired specifications.

Real-Time Monitoring and the End of “One-and-Done” Fixes

The traditional Six Sigma approach to process improvement often follows a linear path: define the problem, collect data, analyze root causes, implement changes, and then control for future variation. However, in many organizations, the “Control” phase is where improvements stagnate. Businesses implement fixes, but without real-time process monitoring, inefficiencies creep back in over time.

Here, too, Process Mining is a game-changer. By continuously monitoring Cp and Cpk scores across different process segments, businesses can detect early signs of performance drift. If a process that once had a Cpk of 1.5 suddenly drops to 1.1, an automated alert can trigger an investigation before inefficiencies escalate. This transforms process improvement from a reactive, episodic exercise into an ongoing, proactive strategy.

In industries like financial services and healthcare, where regulatory compliance and quality control are paramount, this capability is particularly valuable. A hospital, for instance, might need to ensure that patient discharge processes meet strict compliance standards. If Cpk values start declining due to procedural deviations – perhaps a new software update has introduced an extra, unintended step – Process Mining can immediately flag the issue before it affects patient care.

From Historical Analysis to Predictive Process Optimization

The integration of Process Mining with Six Sigma is not just about improving how organizations measure processes – it is about redefining how they optimize them. Traditionally, businesses apply Six Sigma methodologies to historical data, identifying patterns after problems have already occurred. But with the AI-driven advancements in Process Mining, organizations can now forecast process deviations before they happen.

By analyzing vast volumes of historical process data, AI can predict when and where a process is likely to drift out of specification. If past data suggests that a particular supplier frequently causes late deliveries under certain seasonal conditions, predictive analytics can trigger preemptive process adjustments, such as rerouting orders or adjusting inventory buffers before delays occur.

This shift from reactive process improvement to predictive process control is what makes the combination of Process Mining and Six Sigma so powerful. It enables businesses to move beyond periodic performance assessments and embrace a model of continuous, intelligent process optimization.

A New Standard for Process Excellence

The digital transformation of business operations demands a new standard for process excellence – one that is no longer reliant on manual data collection, static reporting, or after-the-fact analysis. The integration of Process Mining with Six Sigma process capability analysis represents the future of operational excellence, offering a level of precision, transparency, and agility that was previously unattainable.

For executives and process improvement leaders, the message is clear: in an era where speed, efficiency, and quality are paramount, organizations that embrace real-time process capability monitoring, AI-driven root cause analysis, and predictive process optimization will gain a decisive competitive edge.

As businesses navigate the complexities of modern operations, the true differentiator will not just be process efficiency – but process intelligence. Those who can see, understand, and optimize their processes in real time will set the standard for the next generation of operational excellence.

Authors

This article was written by Josua Reimold and Jutta Reindler.

Sources

Graafmans, Teun & Turetken, Oktay & Poppelaars, Hans & Fahland, Dirk. (2021). Process Mining for Six Sigma: A Guideline and Tool Support. Business & Information Systems Engineering.

Geffen, F.V., & Niks, R. (2013). Accelerate DMAIC using Process Mining.

Pongboonchai-Empl, T., Antony, J., Garza-Reyes, J. A., Komkowski, T., & Tortorella, G. L. (2023). Integration of Industry 4.0 technologies into Lean Six Sigma DMAIC: a?systematic review. Production Planning & Control, 35(12), 1403–1428. https://doi.org/10.1080/09537287.2023.2188496

Ramires, Francisco & Sampaio, Paulo. (2021). Process mining and lean six sigma: a novel approach to analyze the supply chain quality of a hospital. International Journal of Lean Six Sigma. ahead-of-print.

Six Sigma (2024). Six Sigma Principles. Process Capability Index (Cpk) for Business Success.https://www.6sigma.us/process-improvement/process-capability-index-cpk/