Process mining: Business development through beautiful analytics

#processmining #advancedanalytics #businessanalysis #businessprocess #dataanalytics

Most business activities nowadays run on or at least supported by some sort of information technology applications, the data that is stored in these systems become extradentary valuable for optimizing the business performance and better satisfy customers.?

Analyzing business data?

Business uses various traditional approaches for process analysis and process optimization, the practice usually starts with the development of business process models that describes the business process steps and identify the actors and the systems involved, with the expectation that the actual process will go as planned.

An example of a business process; New bank account opening (simplified)

source : https://www.businessprocessincubator.com/content/is-your-bank-prepared-for-generation-z/

When the designed business process is deployed to support live business operations, it is not unusual to encounter various “Deviations and Gaps” from the planned process, transaction flows might not necessarily be following the “Ideal” process flow as planned, Inefficiencies could be detected and management keeps requiring optimization to the process to make business more efficient and more competitive.

Traditionally, Business and data analysts will start preparing reports and identifying key performance indicators in an effort to measure the efficiency of a business process. They will start exchanging “Subjective opinions” on the possible causes of process and performance deviations. ?

The problem

However, there are many “Challenges” facing the traditional approaches to process analysis,

• Many business processes are not managed entirely within a single system, business users can be jumping between applications, exchanging calls and emails with clients, suppliers and other business partners, manual activities might be required for the business process and the list goes on…
• It is quite hard to reach a consensus about the root causes of inefficiencies if all the available opinions are subjective or not sufficiently supported by data. ??
• Actual business process data is not covered completely within the “operational applications” layer that supports the business process; but there are loads of data that are out of this operational application box, like the mentioned data about the client and supplier interactions or manual activities happening as part of the process. ???????
• The realization of a business process into an actual business case (process instance) can be deviated significantly from the design work developed by the BAs, not only that it can be difficult to identify these “deviations” using conventional approaches in the first place, but also the deviation can be changing over time with to respond to changes in business and market. Examples of these deviations are:?

? o??If the vast majority of cases are going through a certain path, rendering the other paths inefficient or redundant.

o??If many cases the business cases are experiencing repetitive rework looping within a certain path in a way that was not expected during the initial business process design ???

• It is required to get the bigger picture of how the business is performing, pure statistical reports can give false indicators and derive inadequate business decisions, for example, assume that a certain process KPI is indicating the average process completion time is 30 minutes- which is acceptable by business- while the underline analytics show that 50 per cent of the transaction take few seconds to complete with the rest may take several hours- service time that cannot be accepted by business- So high-level reporting can be misleading sometimes.?

Source:?https://axistoconsulting.com/isnt-there-an-opportunity-in-every-challenge-part-4-process-mining/

What is process mining and what does it offer?

Process mining can be considered a special branch of data science, its distinction is combining data mining and process analytics to derive stories and insights from business process data. It leverages a data-driven approach to process optimization, allowing managers to remain objective in their decision-making around resource allocation for existing processes.

Process mining can be used to deliver “Useful insight” into how the business operates, this can be advantageous for organizations of various sizes and industry sectors and it can be applied to any workflow within these organizations, the generated process insights can generate many benefits for an organization such as:

• Process understanding, understanding how the actual business cases workflows differ from the designed business processes
• Identifying the root causes of deviations and inefficacies; studying what delayed, or deviated events share in common is key to identifying the root cause of these nonconformities. It can be surprising how much business impact can be caused by a faulty piece of equipment or an unqualified employee or an unexpected path in the process flow.
• Process optimization and eliminating redundancies, if an organization can identify the deviations and causes of efficiencies, they will be able to do something about it!.
• Customer behaviour analysis, by understanding the business cases, organizations can find more detailed answers to questions like “what” customers are doing to “why” they are doing it. Process mining can provide a better in-depth look into the customers’ journey to help establish patterns and understand behaviour.
• Market change assessment, the typical process might change over time due to a change in market or customer behaviour, it is very beneficial to capture these changes and respond to these changes as early as possible, either by exploiting new opportunities or by mitigating risk. ???
• Process automation support if we understand the business process adequately, we can build and enhance the systems and business practices that
• Process conformance assessment, analyzing and understanding business processes provide information to determine whether a process or service complies with the requirements of a specification, technical standard, contract, or regulations.
• Process standardization and unification, a complete interpretation of the processes allow for identifying any deprecated or unrecommended process flows, eliminating redundant activities and possibly unifying the process across business disciplines and interaction channels. ?
• Supporting business mergers, with the proper understanding of how things are getting done in business entities, a merger of the operations can be more efficient and successful.
• Enhancing customer experience, when the business processes are optimized, the customer is expected to receive better service with no loops and rework activities.
• Optimizing application UX design; it is an interesting aspect, but the user activities on an application can also be treated as processes with associated activities, studying these processes and activities can provide an opportunity to optimize and enhance the design.

The process of the process mining

Modes of process mining?

Process mining can be approached with several “Perspectives” these intentions or can vary from one organization to another, or from one workflow to another:

• The control-flow perspective is about the order of activities, and the goal here is to find the optimal path for performing a process.
• The organizational perspective targets the resources involved in a process, i.e., roles, departments, etc. The general purpose is to come up with an optimal structure for the organizational units.
• The time perspective is focused on the timing and frequency of events.
• The case perspective considers the properties that are inherent in different cases or types of processes. The further analysis defines relations and hidden dependencies between these properties, giving a deeper understanding of the problem and its causes.

(Source : https://www.altexsoft.com/blog/process-mining/)

Data collection and curation?

To fuel the process mining analytics, we will need to capture all the traces of activities with business processes, these traces (commonly called the transaction footprints) to form the source of data that powers the process mining. The captured transaction details are stored on what we call an organization “Event log”.

So, an Event log is predominately constructed by combining information about processes and activities from various operational systems, augmenting the collected information with all available to capture information about activities that take place outside of the operational systems. Standardizing these data into a homogenous business event log.

The Event log needs to cover at least four aspects of information about each activity:

• A way to identify the transaction or process instances (also called a case ID), these identifiers are crucial to identify and document all activities and time stamps that are directly related to each process instance (Case), examples of a case id are “Service ticket number” or “Order number”
• Information about the activities happening (process step or transaction) such as the “break inspections” step in an MOT inspection process or a “Document review” activity in a bank account opening process. ?
• Identifiers for and the objects associated with it (Also called the case object IDs) such as the IDs of the employees who were involved in a specific activity or the Id of the warehouse where a shipping activity is souring from.
• Information about the timestamp when an activity took place ?

Data processing

IEEE designated a “Task Force on Process Mining” to promote the topic, thee taskforce developed the “Process Mining Manifesto”?( https://www.win.tue.nl/ieeetfpm/downloads/Process%20Mining%20Manifesto.pdf ) - an interesting read if you want to get more details on the process mining

The authors of the manifesto identified three categories of process mining techniques.

• Automatic Process discovery is a primary technique and implies extracting and visualizing process models from an event log.
• Conformance checking compares the actual process with a predefined model to discover deviations. So, it’s used to check if the reality conforms to an existing pattern.
• Enhancement?goes beyond measuring and comparing. It’s focused on extending the model with additional information, such as the involved parties, location data, costs, and timing.. these additional data components are fed to the processing engine to identify the possible alterations to the process for optimizing factors like cost and time.

After choosing the desired scope, perspective and technique, process mining technologies are used to apply the processing techniques to the event log data to automatically visualize and reconstruct the actual process flows for all the “Instances” of every process in scope, so that the business owners can get an actual transparent view on how the business processes actually run in their business.

Visualizing and analysing the processes

With the data availed and processing techniques applied, the process mining solutions produce beautiful dashboards illustrating among many other things the overall business process flow, the percentage of cases going through every path and statistics about the cases within any steps.

Filters are provided to limit the visualization to only the cases that satisfy certain criteria such as ordering a specific service or interacting with a specific branch..

The following samples of process mining dashboards speak for themselves, showing the power of process mining to reveal the details of what is actually happing in operations, allowing pinpointing the inefficiencies and identifying the root causes of performance snags.

Image source: www.processmining-software.com

Image Source https://www.process.science/qlik-sense/product