Next article
Process adherence for operational excellence: Closing the gap between theory and practice
When business processes adhere to how they were designed, a company gains the cost savings, agility, governance and quality needed to thrive and compete in their markets. Process Adherence uses process mining to help an organization reconcile their “as-is” process execution with their “should-be” operating model in a way that is seamless, continuous and connected to their strategic goals. This enforces best practices in process execution, keeps undesired deviations at bay and provides an opportunity to evolve existing process best practices into innovative “next practices".
To achieve optimal process adherence, two approaches are used; process compliance and process conformance.
Process Compliance refers to ensuring regulations and industry standards are met. Some of them are optional but they are important quality indicators for any organization (e.g. ISO9001). Others are industry specific (e.g. ANSI product certifications) and some are general laws (e.g. GDPR) or, industry specific laws (e.g. HIPAA). In some cases, non-compliance can result in severe penalties or even imprisonment for leaders.
Other standards might be related to SLAs that companies promise to their customers (e.g. max delivery times) or even internal business rules (e.g. do a credit check in case an order has a volume higher than 100k) and policies (e.g. hotel cost restrictions per country) that are highlighted in employee guidelines. All of these regulations are usually written down in lengthy documents, but it presents a challenge when trying to express the concrete rules through process intelligence software in order to monitor their compliance. Existing concepts are too restrictive and only allow for a small fraction of what would be required, and encoding even those require heavy technical expert knowledge.
Process Conformance ensures that a process is executed in reality as it was designed. Conformance checking is a process adherence technique that compares actual execution data (i.e. the event log) of a process to the should-be model of the same process and highlights the differences. Conformance checking can be used for process compliance however, a key challenge is the lack of flexibility of the designed process model. In reality, there are often many valid ways to execute a process. If all valid behavior were captured in the shuold-be process model, it would become too complex for humans to understand. The work around is dealing with long white lists, which are hard to maintain. In addition, and specifically when modeling and mining projects are driven by different groups of people, there is a strong disconnect between the system generated events (the event log) and the activities captured in the model. There is also a different level of granularity that is very challenging, if not impossible to resolve without significant rework in modeling.
Process Adherence brings the elements of process mining, process modeling, process compliance and conformance together in a powerful way. Working with each piece in isolation puts the onus on the user to find insights, creating a longer time to value. However, because these elements are unified through process adherence, the most relevant insights for the organization are brought to the user. This revolutionary approach takes the complexity out of managing separate disciplines in order to enable the fastest and most seamless value journey possible.
What is the Process Adherence Manager
Process Adherence Manager (PAM) is a solution from Celonis that serves as the backbone to structure and steer processes in a way that aligns with organizational objectives, fosters efficiency and reduces risks. PAM enables users to identify value opportunities by comparing their event data against a “should-be” process model in a guided way. PAM combines the four disciplines of mining, modeling, conformance and compliance analysis into one user interface.
The 4 steps to Process Adherence
Design the should-be: A proper should-be definition consists of two things, (1) a process model and (2) business rules. What is unique about PAM is that you can model in an accelerated fashion by beginning with the insights you glean from process mining to ensure the right level of detail and scope. On top of this, you define behavioral rules for path routing, performance and quality related requirements and all other things that are important to run your business with excellence. With Celonis you can seamlessly use PAM as a baseline to model your process and start into a continuous improvement cycle in Process Designer using AI to accelerate the modeling experience.
Monitor adherence: PAM uses your event data to analyze whether processes run in accordance with the should-be definition. The adherence rate tells you what portion of processes are running as desired. For those that don’t, PAM tells you where they deviate. Going a level deeper, you can see deviations and determine root causes. This provides you with clear and easy to understand insights that show where your processes behave differently than they should.
Assess deviations: PAM evaluates the significance of each deviation in terms of impact on your key performance indicators (KPIs). Those can be general performance metrics such as the process throughput time, on-time-delivery rate, or working capital. This allows you to easily spot the value opportunities that have the biggest impact on your strategic goals.
Take action: PAM uncovers the root causes behind deviations, by telling you what the driving factors, such as a product, a vendor, a sales region, etc. for their occurrence are. Having those highlighted allows you to take targeted action - i.e. by updating your process model in Celonis Process Designer, or using pre-built apps or automatic action triggers. To ensure you are maximizing your return on investments, the Celonis Transformation Hub helps you to track the value generated as well as the adherence trend on your journey.
Adherence Management in Action
There are several scenarios for adherence management across industries that can drive enterprise wide value and help your organization better execute on its strategy. Let’s take a look at some of them.
Manufacturing Line Efficiency (Process: Production Workflow)
Process Model: Predefined production workflow that includes steps for raw material handling, assembly, quality checks, and packaging.
Deviation: Process adherence shows that assembly times vary across shifts. The non-conforming activity involves some workers skipping intermediate quality checks, leading to rework further down the line.
Root Cause Analysis: The skipped quality checks are due to time pressure in high-demand periods, where workers try to speed up the assembly line by bypassing checks, causing higher defect rates.
Update to Process Model: The process model is updated to mandate automated quality checks at critical points, reducing reliance on manual checks and ensuring that no step is skipped under pressure.
Outcome: The automated checks reduce defects by 20%, aligning the actual process with the ideal production model and improving overall line efficiency.
Healthcare: Patient Intake and Treatment (Process: Emergency Room Workflow)
Process Model: Structured patient workflow from triage to diagnosis, treatment, and discharge within a specified 3 hour discharge window.
Deviation: Process adherence reveals that in 30% of cases, discharge times exceed the target due to manual data entry delays. Non-conforming activities include staff manually updating patient records after treatment.
Root Cause Analysis: The delay arises because staff are often overloaded and switch between treating patients and completing paperwork, leading to inconsistent record updates.
Update to Process Model: The model is updated to integrate real-time electronic health records (EHR) systems that automatically update treatment status, allowing for faster patient discharge.
Outcome: The automated system reduces manual entry by 40%, bringing discharge times closer to the target window, closing the gap between actual and ideal performance.
Supply Chain: Order Fulfillment (Process: Supplier Onboarding and Order Processing)
Process Model: Standardized onboarding and order processing workflow with clear timelines and quality control steps.
Deviation: Process adherence shows that 20% of supplier orders miss deadlines. The non-conforming activity is a lack of adherence to the onboarding process, with suppliers starting deliveries without proper timeline agreements.
Root Cause Analysis: New suppliers skip the formal onboarding process, leading to unclear expectations and delivery delays.
Update to Process Model: The updated process mandates that no orders can be processed until suppliers complete onboarding, including agreeing to delivery timelines and quality checks.
Outcome: By enforcing this policy, on-time deliveries improve by 30%, aligning supplier performance with the ideal order fulfillment process.
Financial Services: Loan Approval Process (Process: Loan Application and Review)
Process Model: Standardized loan application workflow with document verification, credit checks, and approval within 48 hours.
Deviation: Process adherence identifies that 20% of loan applications are delayed due to manual document verification. Non-conforming activities include loan officers manually reviewing documents when they should rely on automated verification systems.
Root Cause Analysis: Manual reviews are triggered when the automated system encounters formatting issues with submitted documents, which the officers then review manually, causing delays.
Update to Process Model: The process is updated to include a document formatting validation step that ensures all submissions are in a compatible format before they reach the loan officer.
Outcome: Automation of the validation process reduces manual reviews by 50%, speeding up loan approvals and improving conformance to the ideal 48-hour target.
Retail: Inventory Replenishment (Process: Stock Management and Reordering)
Process Model: Automatic stock management process where reordering is triggered when stock falls below a certain threshold.
Deviation: Process adherence reveals that high-demand products frequently go out of stock, with reorders not triggered in time. The non-conforming activity involves the stock threshold not adjusting for seasonal demand spikes, causing delays.
Root Cause Analysis: The stock threshold for high-demand products is static and doesn’t account for increased sales volume during peak seasons, leading to stockouts.
Update to Process Model: The reordering process is updated to include dynamic thresholds that adjust based on sales trends and forecasts, ensuring orders are triggered earlier for high-demand items.
Outcome: Stockouts are reduced by 40%, improving availability and aligning inventory management with the ideal model.
Process Adherence is a proven way to help close the gap between should-be operating models and as-is execution. With the speed of change and technical disruption continuing to pressure organizations to evolve, an effective method for bridging this gap between theory and practice efficiently is necessary and Process Adherence Manager can be the gateway to your optimal performance.
This article was co-written by Tom Thaler, Principal Product Manager at Celonis, and John Santic, Director of Product Marketing at Celonis.
