Process mining for customer journey mapping
What is business process improvement?
From reducing costs, execution times, and errors, businesses are always seeking ways to improve their operations, otherwise defined as “business process improvement.”
The way a business process is designed and performed determines the quality and efficiency of its service delivery. To improve the output of products and services, organizations must reassess and improve their business process.
A deep dive into business process management
Business process management (BPM) is a discipline that brings together methods to discover, model, analyze, measure, improve, optimize, and automate business processes1. A continuous BPM cycle consists of the following six phases:
- Process definition – the first phase, in which a business problem is formulated. Processes that are associated with the formulated problem are defined together with process performance measures.
- Process discovery – where the current state of each of the defined processes is described. As a result, ‘as-is’ process models are created.
- Process analysis – where issues from the ‘as-is’ process are defined, measured, and prioritized based on the importance and complexity of solving them.
- Process improvement – in which the goal is to define changes that would improve the process and solve key issues. As a result, the ‘to-be’ process model is created.
- Process implementation – in which the changes needed to get to the ‘to-be’ process from ‘as-is’ are made.
- Process monitoring – where the process is being analyzed and evaluated continuously to keep up to date.
The importance of the discovery phase
Each of the described phases of the BPM has its own implementation methods. One of the most important and complex is the process discovery phase, where the ‘as-is’ process models are created. There are several methods for implementing the process discovery phase, such as evidence-based, interview-based, and workshop-based discoveries. There is also customer journey mapping, which combines several different discovery methods.
What is customer journey mapping, and what are the applications?
Customer journey mapping (CJM) is an excellent way to perform process discovery and deep-dive into the business’ process.
Traditionally considered a tool for visualization and analysis of touchpoints between the customer and product, the aim of creating a CJM is to structure customers’ behaviors and define the most effective methods of engaging them.
CJM helps organizations understand what their customers think, feel, see, and hear. By investigating touchpoints, an analyst can discover customer ‘pain’ and then do everything possible to remove or reduce it.
There are several types of customer journey maps, including:
- The ‘day in the life’ model to explore a new product or service idea
- The ‘current state’ model to create the ‘as-is’ model and enhance an existing product or service
- The ‘future state’ model to create the desired ‘to-be’ model for the process improvement phase
- The ‘service blueprint’ model to define the root causes of current customer journeys and is useful for the process analysis phase
Components of a customer journey map
There are seven main components of a CJM:
- Customer profile or personas: describes a group of customers using the product or service.
- Stage: the step of the investigated process.
- Goals: describes goals that customers want to achieve through each stage.
- Touchpoints and channels: the most significant component of a CJM – shows all points of interaction between the customer and the company, as well as channels where this interaction occurs (e.g., website, app, delivery service, portal).
- Pains and gains: challenges and benefits that the customer encounters.
- Empathy and emotions: how the customer feels, thinks, hears, sees, says, and does at each stage of the CJM.
- Opportunities: a description of ideas that can help to improve the customer experience.
Is a customer journey map a must-use tool?
Each customer has their own unique path. However, by using CJM, an analyst can only determine a typical customer’s behavior. It is hard to assess how close this typical path is to reality and what percentage of all customers fall into the typical path. It is often the case that an actual customer journey is quite different from the expected customer journey, as is visually represented below:
The expected customer journey
The actual customer journey
Moreover, it is worth remembering that creating an accurate customer journey map is no easy feat and can be quite time-consuming. Describing the business process in detail is at least a time-consuming task. However, with the help of other techniques, such as process mining, customer journeys can be created more efficiently and accurately.
How can you use process mining to create customer journey maps?
Process mining is a discipline that connects computational intelligence and machine learning with process modeling and analysis. It is a set of techniques, tools, and methods for real process discovery, monitoring, and improvement, that provides an opportunity to evaluate and control the business process by investigating deviations and bottlenecks to address or eliminate later2. Process mining combines ideas from both process science as well as data science:
How to apply process mining techniques to customer journey mapping?
The approach of using process mining techniques for customer journey mapping creation is based on the concept of continuous improvement of business processes.
There are certain business processes where a customer is engaged. The customer interacts with the information system that supports this business process. Any interaction between the customer and the information system is automatically recorded in the event log. An event log as a source of data conceals a lot of insights. Fundamental techniques from process mining and machine learning can be used to gather and implement insights to improve the business process.
How can a customer journey be reconstructed from an event log?
More and more customer interactions with the product are happening through different information systems (e.g., via phone, tablet, or computer), meaning a large portion of touchpoints are now digital.
The classic customer journey consists of seven main stages:
- The emergence of the need.
- Creation of interest regarding a product, preliminary product research.
- Engaging in interaction with the product.
- Purchasing.
- Usage of the product.
- After-sales service quality assessment regarding the product.
- The product recommendation to friends and positive reviews on social media.
Part of the traditional customer journey stages can be available for analysis using the proposed approach. Usually, stages 3, 4, and 6 are made by the customer using a single information system. The information system automatically creates event logs, which provide huge opportunities for analysis. As a result of the analysis, a CJM could be created that shows the customer’s interaction with the product, with additional indicators like time and frequencies. Below is an example of a customer journey obtained using an event log.
The journey represents activities performed by a particular person, such as Browse services, Examine service details, Browse parameters, etc., that correspond to stages 3, 4, 6 of a classic customer journey.
Challenges of a process mining approach to customer journey mapping
Unfortunately, not all stages of the classical customer journey are digitalized and available for analysis, meaning the available data may only display part of the customer journey.
Additionally, data available from an event log is often confusing, not labeled, and hard to understand, requiring a lot of effort to process and analyze. There are no unified algorithms to build CJM from event logs, and it is important to remember that each case and business process is unique and requires an individualized solution.
While there is no catch-all solution for these difficulties, experience shows that the most successful projects are those conducted in collaboration with both business experts and PM/ML specialists, enabling the expert in business process to assist the PM/ML specialist, who is equipped to solve riddles in the data.
References
- Marlon Dumas, Marcello La Rosa, Jan Mendling, Hajo A. Reijers. Fundamentals of Business Process Management, Second Edition. Springer, 2018.
- Wil van der Aalst. Process Mining: Data Science in Action, Second Edition. Springer, 2018.
Additional insights
Why process mining is the top skill to learn for business analysts in 2022
Metrics that matter: How UX measurements drive product success
Are you set up to successfully compete with experiences?
