Process Mining and Task Mining: The Perfect Combination for Successful RPA

Process mining and task mining are vital elements of any intelligent automation project. However, there’s a great deal of confusion about the tools and whether they can be used interchangeably or are actually unrelated.

The answer is that they serve different purposes and applications for businesses. However, when used together in a complementary way, process mining and task mining serve to help transform the way businesses operate. Let’s define what the tools are and the crucial differences between them.

What’s the difference between process mining and task mining?

As their names suggest, process mining and task mining are all about discovery. If an auto manufacturer wants to improve the efficiency and cost-effectiveness of how it makes a new vehicle, it will review how well its processes take a car from initial design to completion.

The manufacturer will also look at the steps a person or machine uses to undertake specific tasks, including crafting sheet metal, fixing a part or building the transmission.

In a broader business sense then, process mining examines how well a system flows between people or machines. Task mining uncovers how efficiently individual parts of a process are being managed. Therefore, process mining and task mining are complementary in that they both consider how to improve an overall system but from different perspectives.

Process mining in detail

Process mining is a set of techniques that organizations use to understand processes and sub-processes, monitor key performance indicators (KPIs) and identify potential problems. It looks at complete end-to-end processes, including elements that run across different IT systems and teams. That means it’s possible to look at how each step relates to the other and how they contribute to a specific outcome or set of outcomes.

For example, when undertaking process mining, organizations may identify missing or duplicate steps or bottlenecks and root cause problems that make the process inefficient.

A good example is staff onboarding, which involves multiple steps and teams across the organization, as well as various outputs that must happen before a new team member can start work. It could involve checking professional qualifications, producing security passes, creating an IT system log-in, or setting up payroll, pension and benefits.

Process mining helps organizations visualize each stage of onboarding — which steps must happen when, who is responsible for each task, and where there are interdependencies. For example, a new team member needs to provide their home address and banking details before payroll can be set up. IT needs to know which apps and systems the new joiner needs access to before they can provide an email address and password.

This visualization of what processes have been discovered shows the real process flow with all the variations, cases, and paths based on event logs. These are data sources that have been pulled from core systems that display activities — what it is that was performed and when. Data is typically taken from customer relationship management (CRM), enterprise resource planning (ERP) and other IT systems.

Task mining in detail

Task mining focuses on the smaller components of a process or subprocess and is usually performed on the employee’s computer. It drills down into the granular clicks of an individual or compares clicks of multiple users to see how different people completing the task compare.

With task mining, organizations can see frequency, complexity and cycle times by recording keystrokes and clicks, as well as copying or pasting activity to build a picture of how a task is typically undertaken.

The data that is produced with task mining helps organizations understand the steps needed to perform a task, as well as variations between different users. This provides an accurate picture of how tasks can be optimized and improved, where they can be automated and what improvements this could bring for the employee as well as customers.

Tasks might involve copying data from a CRM or finance system, moving between multiple systems to identify data or analyzing incoming information using toolsets. Blue Prism Process Intelligence Powered by ABBYY Timeline enables organizations to undertake four key steps:

Task Recording

• Records user interactions to collect data
• Deploys on an employee’s desktop
• Takes screenshots and captures information at intervals which are sent back to the software

Task Analysis

• Identifies patterns of user interactions
• Detects variables in task execution

Context Recognition

• Detects all user interaction data during the recording process and is converted into structured data using OCR technology

Advanced Custom Recording

• Takes advanced custom recordings to get more detail
• Can be approached in different ways in terms of steps performed, time spent and time lost

Task Mining Versus Process Mining Process Mining Task Mining What can be achieved Understand the flows involved in processes and how to improve them with automation Analyze tasks undertaken by users and determine where they can be simplified and automated Data source Core IT systems End users’ PCs End results More efficient, automated end-to-end processes Simpler, more effective tasks undertaken by users

How Process Intelligence combines process mining and task mining

Blue Prism Process Intelligence Powered by ABBYY Timeline combines process mining, task mining and business intelligence-like metrics to provide detailed insights into processes and tasks across the business. It adds significant value in three main ways:

• Creates a complete history of every process iteration from beginning to end, even when some steps are performed using multiple systems. Recorded as timelines, processes and tasks can be compared, filtered, searched and aggregated to identify inefficiencies and potential for improvement.
  
  
• Monitors every iteration and alerts process owners about any deviation that could cause delays. It enables continuous improvement and continues to deliver return on investment (ROI) as businesses operate, not just at the initial stages.
  
  
• Allows organizations to define process rules, then trains the system to identify problems and alert the business. This approach can extend to alerting a human to step in and fix the issue quickly and efficiently.

Why Process Intelligence is the perfect blend for intelligent automation

Organizations that use only process automation or task automation tools are missing an opportunity to automate and digitize at the same time. To succeed with digital transformation, organizations need visibility into how people, processes and tasks are operating within their businesses, where the bottlenecks are and how well automated processes are performing in practice.

Achieving results quickly using traditional methods, such as interviewing process owners and those who undertake tasks, is impossible. Not only is it a time-consuming approach, but it can introduce bias. With Process Intelligence, organizations can rely on data-driven, real-time insights that take them through to the next level of intelligent automation.