Editor's note: this is a guest post. Views expressed in this article are the author's own views and not necessarily representative of UiPath.
Evolving customer expectations, intensifying competition, changing market dynamics, and expanding regulatory coverage and stringency have necessitated firms to build lean, resilient, and agile business models. COVID-19 has further exposed survival issues and business continuity risks that exist in legacy business models. Consequently, enterprises across industries are looking to accelerate the shift to digital, automated, and intelligent business processes.
While many enterprises have successfully piloted initiatives for low-complexity use cases, they continue to struggle with scaling these projects to enterprise-wide programs. Consequently, the value realized from these initiatives remains much lower than their overall potential.
While enterprises face several challenges in their automation and digital transformation journeys, identifying and sustaining a healthy pipeline of process optimization and automation opportunities remain as key barriers to scale and value realization. Without a healthy pipeline, organizations merely have big goals and the urgency to change but no clear roadmap for success.
Challenges in creating a healthy automation pipeline
A healthy automation pipeline is much more than having a bunch of use cases and opportunities and is comprised of four key requirements:
Fact-based approach to discover as-is processes at scale across both macro (i.e., broader organizational workflows) and micro (i.e., task or end-user activity) levels
Ability to identify process optimization and automation potential in an integrated manner
Holistic view of the ROI (i.e., going beyond cost savings) while prioritizing the process optimization and automation opportunities
Ability to continuously monitor the actual impact and ROI of the implemented initiatives and automations. It enables a continuous feedback mechanism that helps validate and refine the transformation pipeline and roadmap in an agile manner
We find that organizations are unable to create a healthy transformation pipeline due to a heavy reliance on manual, interview-based techniques. This commonly used approach is plagued with various limitations and challenges such as it being extremely time- and resource-intensive, subjective, perception-based, potentially biased, unable to capture all process variants, and resistance from stakeholders in providing necessary details.
Need for process mining
Process mining helps overcome the limitations of manual techniques and challenges in building a healthy pipeline since it:
Follows a fact-based approach
Provides greater depth and breadth of information
Is easier to scale and more cost-effective than manual techniques
Enables continuous monitoring of process performance and ROI
Process mining technologies can be classified into three categories: classic process mining, desktop process mining (also called task mining), and hybrid process mining.
The classification of these tools is driven by the type of process data collected, the insight derived from the collected data, and the associated use cases the insights support.
Classic process mining
These solutions capture process-related information from event logs generated by enterprise systems such as enterprise resource planning (ERP), customer relationship management (CRM), and supply chain management (SCM). The captured details are used to reconstruct as-is processes and help visualize process flows, step repetitions, and variations at a macro level. These solutions further help monitor process performance metrics and key performance indicators (KPIs) on an ongoing basis and identify process improvement opportunities.
Desktop process mining/task mining
These solutions capture process-related information from end-user interactions with workplace applications. This includes keystrokes, mouse clicks, application object IDs, screen information, and other potential system-level activities. This granular data gives insight into the tasks and activities involved in executing a process at a micro level. The insights from desktop process mining help identify task-level automation opportunities, generate process documentation, and monitor processes and employee productivity.
Hybrid process mining
The demand for a unified and holistic view of as-is processes is fueling the need for a hybrid process mining solution that leverages both event and user interface (UI) logs to provide the best of both worlds. While such solutions are still in their early stages of maturity, they are expected to evolve rapidly.
Applications of process mining
Process mining plays a critical role across all stages of enterprises’ digital transformation lifecycles to accelerate time-to-value realization, ROI, and scale. Process mining can help provide the following benefits:
Discover: uncovers or validates existing process flows along with associated deviations, exceptions, variances, and essential step information (such as time, cost, volume, and frequency). They can also be used to create or update process documentation, which can be leveraged for use cases such as training new employees and handing over processes in outsourcing scenarios.
Optimize: compares as-is processes with defined standard operating procedures (SOPs) for conformance checks and analyzing deviations in terms of path, time, and costs. It helps streamline and reengineer processes by identifying exceptions, blind spots, and process vulnerabilities as well as by rationalizing the number of variants identified. Process mining also helps derive insights into employee collaboration, which can be used for better resource allocation and delegation of work.
Automate: captures extensive process information such as volume, costs, time, and frequency of process paths and steps. This information is analyzed by the process mining tool using appropriate frameworks to identify automation opportunities and eliminates the reliance on hunches and opinions.
Evaluate: simulates various process optimization and automation scenarios to predict the ROI and create or validate the business case. Common simulation analysis approaches involve configuring what-if scenarios by defining certain attributes, using process filters to compare steps, and examining the impact on relevant KPIs such as throughput time and rework. These simulations minimize the risk of implementing improvement projects without the knowledge of how they will impact real-time operations.
Execute: reduces the execution gaps by automatically triggering actions based on insights generated. Helping act on the insights generated is becoming a quintessential part of process mining, which is also driving its integration with various complementary solutions such as robotic process automation (RPA), workflow automation, and case management. Consequently, a wide range of actions can get triggered depending upon the use case, including:
Alerts and notifications about events requiring attention (e.g., KPI breaches, service-level agreement violations, etc.) via email and dashboard displays
Assigning tasks to relevant users or creating tickets
Triggering RPA bots and automations to carry out certain tasks (e.g., high-severity tasks when a user fails to act within a specified duration)
Monitor: enables an organization to monitor process performance on an ongoing basis and create a roadmap for continuous process optimization and automation. Monitoring processes in near real-time helps enterprises detect bottlenecks and predict any challenges in meeting critical service-level agreements (SLAs) or any potential KPI breaches, and plan for remedies.
Understanding the process mining journey
Adoption of process mining can be driven by ad hoc needs to analyze a few processes or as part of an organization-wide strategy to transform operations. In both cases, enterprises can break down their process mining journeys into these five distinct steps for greater success and value realization:
Understand the current state: enterprises need to understand the current state of their process mining capabilities and outcomes as well as possible outcomes achievable
Create a business case for the desired outcome: as a next step, enterprises should identify processes suitable for process mining implementation, create a business case for the desired outcomes, and refine the target outcome state if the business case does not stand
Determine the capability target state: having determined achievable outcomes, enterprises should map out corresponding capability requirements to achieve them. Everest Group evaluates process mining capabilities across more than 25 key elements of enterprises’ process mining journeys and four maturity levels. Organizations can undergo multiple sprints of varying durations to move from one state to another. They can be at different levels of maturity across their capability elements
Identify all determinants and map the path: given the same current and target states, the process mining journey for different enterprises could take different routes depending upon environmental factors such as organizational structure, people/process-centricity, initiating stakeholders, risk appetite, sensitivity to change, and technology savviness
Execute against the mapped path: having mapped the best-fit execution path, enterprises could leverage various best practice frameworks and tools for successful execution
The process mining market continues to grow rapidly, driven by increased awareness and maturity of process mining solutions. Enterprises across geographies and industries have started to realize their role in accelerating automation and digital transformation journeys as well as fueling continuous process monitoring, which is critical to ensuring business agility and resilience. Moreover, its integration with automation solutions has expanded its role in driving fact-based and outcome-driven actions.
Everest Group’s Process Mining Playbook empowers enterprises at various stages of their process automation and transformation journeys with insights, methodologies, and practical advice to achieve best-in-class outcomes from process mining.
This is a companion discussion topic for the original entry at https://www.uipath.com/blog/automation/process-mining-key-continuous-automation-value-realization