Skip to main content Skip to footer

Process automation is all the rage as companies seek to drive out cost, operate with greater speed and free employees to focus on more high-value tasks. But in our experience, processes as they’re performed day to day are often run more on tribal knowledge than on codified, documented directions and may not be fully known even by the teams that created or inherited them. Ineffectively designed processes can slow business performance and hurt productivity.

We recommend starting the redesign with a clear snapshot of the “as is” process and the application of tools to assess how key processes are performing. That’s where process mining comes in. An insight-driven approach, process mining pulls event-related data from customer relationship management (CRM), enterprise resource planning (ERP) and other applications to form a model of the way a business process actually, versus theoretically, functions within the organization. By capturing how work happens, including who did it, how long it takes and any deviations from average, process mining ruthlessly exposes bottlenecks, non-compliant paths and other impediments often unseen.

Applied properly, process mining precedes intelligent process automation (IPA) and robotic process automation (RPA) initiatives; process mining suites can be used to create automation templates and gauge the health of existing processes — identifying which are efficient and which require an overhaul.

Given the value of automation, it is no surprise that companies are beginning to invest in process mining to optimize how work gets done. Since process mining is relatively young, we field many questions about prerequisites and benefits. We’ll start with where to start and follow with examples for what can be achieved.

Getting started — a checklist

When preparing for a process mining initiative, carefully consider what to tackle first. While some companies understandably begin with a small proof-of-concept project, we recommend instead focusing on processes that require improvement and have significant potential impact. Order-to-cash or procure-to-pay are examples. Improving them can not only drive out cost but also improve productivity and efficiency.

With many organizations investigating this emerging technology, we’ve used our experience to develop a checklist to help simplify a path forward:

Identify business goals

  • What are known pain points/problem areas for the process?
  • What are primary key performance indicators (KPIs) for both operations and business outcomes?
  • Are there compliance or control concerns related to the process?
  • What are the top two or three desired outcomes from process mining?

Determine a timeframe for data analysis

  • How many instances are processed annually by this process?
  • What is the end-to-end cycle time for the typical case?
  • Have there been any material changes to the process execution in the past year?

Identify potential benefits

  • How many full-time employees or equivalents are executing the process (with breakouts by location)?
  • Aside from operating expenses, are there expenses (such as penalties, discounts or deductions) or revenue leakages (such as customer/application losses or reimbursements) from the process?

Identify key stakeholders

  • Who are the business owners, process subject matter experts (SMEs), process excellence resources, source data SMEs, and IT resources (to support process mining application deployment)?
  • Who will be responsible for implementing any identified opportunities?

Identify data sources for the process

  • What are the data sources for workflow/event log data, and how will this data be fed into the process mining tool?

The following figure shows a typical timeline.


What does this look like in practice? We’ve worked with clients worldwide on process mining projects; here are some examples of client work that shed light on real-world impact.

Cleaning up a claims process

For a mortgage lender client’s collateral management claims process, anecdotal evidence suggested there was excessive rework, unnecessary employee effort and non-conforming process variations. Worker productivity was highly variable, and a significant percentage of claims were charged off, rejected or required resubmission.

Our process mining/analytics included:

  • Ingesting four months of case workflow process logs

  • Cleansing and standardizing source data and addressing data inconsistencies

  • Identifying process bottlenecks, variations and associated performance metrics such as cycle time and claim-rejection rate

  • Developing a model to predict the probability of expense rejection to enable workflow prioritization and proactive handling

  • Creating a business intelligence (BI) dashboard for data visualization, graphical presentation and benefit calculation


The outcome included improved workflow processing, resulting in a $3.8 million improvement in cash utilization due to a reduction in the claims rejection rate, and improved customer satisfaction.

Dramatic reduction in cycle time

We recently helped a large healthcare provider identify solutions for the higher-than-normal cycle times required for configuring member benefits. To address the issue, the company sought to identify levers to drastically reduce the baseline cycle time of 26 days while ensuring critical access for all members.

We assessed their operations using process mining to run what-if scenarios and simulations that identified specific opportunities for change, such as process bottlenecks. We identified ways to improve the processes and reduce cycle times. For example, one improvement automatically routed cases of family members with disabled dependents to an agent with specific knowledge in that area to ensure more effective handling, improving customer service.

Using time-and-motion studies, in which screen activity was recorded, the client’s business process team identified thousands of the same clicks being made each day — identifying a manual, necessary process that added time to an already convoluted system.

By applying artificial intelligence to these time-consuming processes, we helped the client reduce cycle time by 70% — resulting in a target of under seven days.

The changes introduced by process mining improved customer experience, reduced human error and created processes that are more efficient and assured. The business process optimization team was able to redeploy employees to focus on more complex, challenging projects that directly created value for the business.

Uncovering wealth

A wealth management firm sought to address issues in its treasury disbursement process:

  • Lack of visibility — no systematic understanding of process bottlenecks and other pain points

  • Inefficient use of resources due to firefighting efforts such as researching case status and exception processing on escalated cases

  • A delay in the completion of cases

  • Customer frustration over delays

Pulling data from a Pega workflow log and other sources, we applied process mining to create an operational dashboard that provides detailed insights on all cases (status, progress, ETA) and predicts completion time for each case, enabling proactive customer management.

The result: A 90% productivity improvement with near 100% compliance with disbursement service level agreements (SLA).

Process clarity: Accelerating business results

The relentless pace of change is here to stay. Process mining, especially when paired with automation and analytics, will help enterprises respond more nimbly to market opportunities. Augmenting human experience with new technologies such as process mining will enable employees to apply precious time to higher value work such as strategic planning, serving customers and innovating new products.

To learn more, visit the Intelligent Process Automation section of our website or contact us.