Enterprises rely on processes, like chefs rely on recipes. However, no matter how much detail is provided, humans are not robots and processes do not guarantee instructions are carried out accurately and consistently. Processes, like recipes, are not static either. They evolve over time as suppliers or regulations change. Quickly, processes can fade or alter, which can leave an organization at risk. It is therefore essential to put some guardrails around processes and to check the health of these processes. Efficiency and effectiveness depend on processes, and they need to be updated consistently to ensure operationalization.
This article will share more detail around process mining and how when combined with Business Process Management (BPM), this powerful pairing can optimally improve processes – and strengthen operations. It also includes use cases in how to apply at any organization.
What is the definition of process mining?
Process mining is an approach that analyzes data from IT systems to gain objective insights and uncover hidden problems. This approach is conducted using software that combines data science and process management to discover, check conformance, and enhance actual processes.
- Discover: The first step is to discover the inefficiencies. The process mining software does this by extracting event logs from information systems such as Customer Relationship Management (CRM) and Enterprise Resource Planning (ERP) tools.
- Check conformance: Next, the process mining tool will review the event logs and turn it into data to identify trends and patterns, and automatically create an "as-is" process model that reflects what really happens in operations.
- Enhance: Lastly, the software will identify precisely where the bottlenecks and inefficiencies are occurring that are causing delays and suggest potential reworks. One can also check conformance by comparing actual processes with BPMN models and optimize them.
How does process mining work
As mentioned earlier, process mining software works by extracting event logs (contained in information systems) to construct processes based on three main attributes:
- Case ID: this allows the software to link back an event to a person or a case
- Timestamp: this datapoint records the time of execution to help sort a sequence of events
- Activity: this attribute corresponds to an activity that was executed in the process
These attributes provide enough detail for the software to identify what stage the process is in. In the example to the right, the phases are “Registered,” “Completed,” “In progress,” or “At specialist.” When an activity is updated for a particular case, the time is also recorded. When looking at the different activities in a macro view, it is possible to construct a process that depicts how cases are actually handled.
Using the event logs and turning it into data, process mining software can identify deviations, bottlenecks, and waste, or investigate their causes and show ways to improve KPIs such as time, cost, quality, or risk. The result of an analysis may be to add more resources at peak times, or to build in support at times where bottlenecks tend to occur.
Use Process Mining with Business Process Management (BPM) to optimize efficiency
Whereas process mining uses data to identify where specific improvements can be made, Business Process Management (BPM) is a discipline focused on aligning processes with business goals. The essential difference between process mining and BPM lies within the outcome. Process mining provides the “as-is” process, while BPM gives a map of the ideal process, also known as “to-be” process. Together these two activities are a power pairing as it optimizes efficiency, ensures customer satisfaction, and drives operational resiliency.
There are five steps to implement this power pairing and ensure processes are accurate, efficient, and strong:
1. Conduct process mining to identify “as-is” process.
In this step, one uses process mining tools that can automatically identify inefficiencies hidden in event logs and answer questions like:
- What is causing delays in my process
- Are people properly trained in doing their tasks?
- Why are some decisions stuck?
Artificial intelligence is a big help in this task. Using machine learning algorithms, process mining tools can compute a combination of different attributes and perform root-cause analysis that helps to identify operational inefficiencies. Root cause analysis is a statistical view that compares a set of data with a reference set of data and analyzes the frequency of an attribute between the two sets of data.
2. Check conformance to see if the “as-is” process matches what was initially designed.
This step allows process analysts to identify gaps in actual processes and deviations from the norm. Some examples include:
- It is common that branches or sites perform the same process differently. By comparing the logs of the different regions, it is easier to make sure that the process is executed in a standardized manner in every branch. This is particularly essential to ensure that the same service is provided everywhere to customers.
- It might also happen that some process tasks are skipped during the execution of the process. Process mining also detects reworks or process loops. One of the reasons of reworks is negligence. For example, invoices that have been incorrectly paid need to be paid a second time or errors that occur when delivering good force agents to perform the second delivery.
Process mining can help uncover all these issues and, importantly, understand the reasons. With real data coming from information systems, compliance initiatives are also easier to audit by verifying that the process being performed actually complies with regulations and policies. Additionally, some business rules can be implemented, so that you are automatically alerted when operational KPIs and standards are no longer met.
3. Use process simulation to review different scenarios to identify improvements.
In this step, it is important to test processes by simulating multiple scenarios (based on time, costs, and resources) to understand:
- What is the impact of adding an extra resource to complete the process?
- What if the allocated time to perform a task is reduced?
- What is the impact on labor costs?
Process simulation enables analysts to simulate different parameters based on statistical models:
- Waiting time of a task: In real life, a task is not executed immediately. For example, for a customer support center, analysts can model that one case is created every 10 minutes on average with an exponential distribution.
- Processing time: This represents the duration of an activity. It may follow a normal distribution with a mean of 15 min and a standard deviation of 4 min.
- Conditional branch (Gateways): These are modeled by defining the percentage of time the conditional branch is taken when the corresponding decision gateway is reached.
- Resource pool: This indicates who is responsible for performing each activity in the process model.
- Timetable for each resource pool: This shows the time periods during which a resource is available to perform activities.
By combining all these parameters, analysts can perform “what-if” scenario analysis, compare different scenarios, and implement the scenario that maximizes operational efficiency.
4. Apply an outside-in approach using a customer journey map to identify the most critical milestones in customer satisfaction.
In today’s age, it is imperative to have an outside-in approach that shows what it is like to interact with your product from the customers’ point of view. To this end, in this step, organizations should use customer journey maps to identify touchpoints (aka interactions) where a customer interacts with the organization. Every touchpoint is rated according to the impact on customer satisfaction, which helps to easily identify and prioritize which touchpoint/s to improve.
Applying this customer satisfaction lens after the first three steps can help prioritize where and when updates need to be made. Let’s face it, if this hasn’t been done before, an organization will likely have many processes to refine and update. Prioritization ensures the business can continue to thrive and grow while also repairing the foundational processes.
5. Strengthen your processes by mapping risks and controls to improve operations.
Lastly, it would be a pity if organizations have come this far (through the first four steps) and not do some housekeeping to strengthen overall operations and apply risk and controls to each process.
Having a view of the risks within business processes enables analysts to improve them and mitigate risks. Using process mining, analysts can gain visibility into why some risks appear in some places.
When to use process mining with business process management?
Process mining technology is particularly effective to improve processes in these use cases organizations have to tackle:>
- Procure-to-Pay - Procure-to-Pay enables the integration of the purchasing department with the accounts payable department. It involves four key stages: supply management, purchase order, receiving, invoicing and payment. In this process, Process mining can help eliminate rework and process changes, automate PO closure and low value invoices processing, and optimize early and late payments.
- Order-to-Cash - Order to cash is the business process for receiving and processing customer orders. Process mining can help standardize the process and minimize PO changes to facilitate on-time delivery and speed up cash collection.
- Customer Service Desk - A customer service desk provides the customer or end user with information and support related to products and services. Process mining helps organizations standardize their assistance process, as well as remove reworks and eliminate pending cases. It also analyzes satisfaction in connection to performance.
- Auditing & Compliance - Process mining helps validate or audit whether actual operations are in conformance with defined operations. More accurate assessments of process deviations and compliance issues, such as segregation of duties, help manage these risks and communicate findings in an audit report.
- Digital transformation - Process mining provides real-time information on how processes perform and how they can be improved using KPIs. Organizations can quickly adapt by continuously monitoring mining data through dashboards and alerts.
- Opportunity for automation (RPA) - It helps organization discover and assess the opportunities for Robotic Process Automation (RPA), by delivering the actual operational data before running into automation.
Want to learn more? Please download our ebook on Optimize your business operations using process mining.