Get the most out of process mining: 3 keys

Organizational preparedness, thoughtful governance, and quality lead to ongoing success.

You've just placed an order for those shoes that will go perfectly with the outfit for your event. You wait with anticipation for the delivery—which fails to come on the promised day. When the too-late shoes finally arrive, you sigh and return them.

You're in good company; 43% of one of our client’s returns were due to late or lost deliveries. The lack of timely tracking was an astronomical cost burden.

But process mining enabled the client to know, track and reduce the actual cost of processing returns in warehouses, anticipate return peaks in real time, and mitigate their impact through capacity planning. Carrier spend was reduced through timely and transparent tracking, which enabled the client to hold the carrier accountable and improve the real cost of a return.

Every organization has inefficiencies in processes that limit its ability to serve customers, shareholders and employees. Disparate application ecosystems and challenges around data management make optimizing processes time consuming and costly.

As businesses seek to address this headache, process mining has enjoyed growing adoption as a technique for discovering, monitoring and improving processes. By extracting data from event logs, process mining can identify inefficiencies, bottlenecks, deviations, and non-standard ways of operation. The insights gained enable organizations to accelerate the remediation of process gaps, increase productivity, and improve experiences for employees and customers alike.

Moreover, with the advent of generative AI, process mining is on the cusp of being a powerhouse for process optimization—it will accelerate data cleansing and preparation, provide more comprehensive insights, and enable real-time analysis.

We’ve listened to our clients and can share three key considerations to help businesses get the most value from process mining.

1.    Build the right team and governance

Process mining success requires close, ongoing collaboration between multiple stakeholders: business, technology, process design and data expertise.

To get a process mining initiative started, business leaders and subject matter experts (SMEs) identify areas ripe for optimization (typically those with high costs, low quality, or significant customer impact—and often those that tie directly to strategic objectives). They work closely with process design experts to gain a holistic view of the “as is” process. Business SMEs ensure the process mining output is a digital twin of the process. These SMEs should also assure regulatory compliance.

Process design experts are needed to pinpoint areas to investigate, analyze mining results, and recommend and design the “to be” state.

Data and technology experts ensure data availability, quality, and connectivity. Technologists ensure the platform is stable and the process mining tools are leveraged in alignment with best practices.

We stress to clients that process mining is a discovery approach—alone, it will not create value. This underscores the importance of an executive sponsor and governance capability. A good governance team ensures the right solutions are prioritized and implemented to achieve the greatest impact.

Beyond prioritization, governance teams ensure there are clear baseline and target values. They hold the process mining team accountable for measuring, tracking and reporting outcomes. Feedback to the governing body is part of a continuous improvement loop that enables course correction and better decision making.

2.    Data readiness

Process mining success hinges on data quality, availability, and curation. Generative AI is a welcome accelerator in process mining. With its ability to analyze and summarize vast amounts of data, it alleviates much of the previously manual work around data cleaning and preparation. Importantly, it enhances the ability to deal with unstructured data. Generative AI, however, is an enhancer to human efforts. Business and process mining SMEs must validate any data cleansing and insights provided by generative AI.

Multi-department processes are complex; to ensure success, process mining and business SMEs must collaborate even before discovery begins. Planning and source system validation includes canonization of a common process flow, creation of a data sourcing strategy, and creation of a data pipeline. Defining a common process flow and sourcing strategy ensures the right systems are targeted.

With data sufficiency established, the focus is on the curation of a data pipeline. Complex environments require a combination of data acquisition methods, such as direct connectors to source systems, ingestion of data from data lakes, and topic subscriptions.

With a data pipeline established, the team can commence with process discovery. Almost all processes have non-system activities, deviations, and other inefficiencies. Business SMEs and process design experts should work together to validate digital twins and identify gaps.

A word about generative AI: It provides a quantum leap in enabling faster and more comprehensive insights. With its ability to quickly understand complex patterns, it enables real-time analysis of process mining insights. Moreover, its predictive capabilities provide proactive analysis and suggestions.

But with or without generative AI, a combination of business and process mining expertise is needed to validate the digital twin. Avoiding false positives and having insight into the impact of errant user behavior enables the team gain quality insights from which practical recommendations can be created.

3.    Organizational preparedness

Process mining is not simply a project. It’s a set of skills, tools and initiatives that support an organization’s data-driven and continuous improvement mindset. Business team members adopt new responsibilities not just during a given project, but on an ongoing basis. Investing in upskilling around process mining tools and techniques and identifying insights from ongoing process analysis is key to gaining long-term value.

Process design experts, business analysts and business intelligence teams should undergo upskilling in process mining analysis techniques. And architects and platform administrators should receive specific training in supporting the platform. It’s often beneficial to pair these resources with third-party experts who can provide mentoring.

Even after a process has been optimized, business users will continue to use process mining tools to create operational dashboards that measure and publish metrics.

Celebrating the success of these initiatives is also important. Early business users should serve as torch bearers, spreading success stories and coaching other team members.

Process mining can unlock human potential by accelerating process optimization. Generative AI promises to accelerate delivery of value through data readiness and analysis. Both process mining and generative AI can support human team members—but cannot replace their business and design expertise.

Success is most likely when organizations include process mining in their data-driven continuous improvement approaches. Senior leadership, cross-functional collaboration, and upskilling are all critical to ongoing success.