Skip to main content

How Does Process Mining Work? Plus 7 Misconceptions Debunked

Kerri Hale, Senior Product Marketing Manager, Process Mining
November 30, 2022

In the past, the process improvement aspect of business process management relied on lengthy and time-consuming research. Process managers held discussions with stakeholders, made guesstimates about how or why a process underperformed, and spent days, sometimes weeks testing enhancements to see if they could observe a difference.

Now, businesses have access to a data-driven process optimization tool: process mining.

How process mining works.

Process mining uses data from system event logs to build a visual representation of a business process. With a transparent view into workflows across systems, along with any process variants, you can quickly and easily identify process inefficiencies and areas ripe for optimization.

Let’s break that description down to understand how organizations can benefit from process mining to understand, monitor, and improve business processes.

Uses data from system event logs: An event log is a record of time-stamped activities produced by a business system. A process mining tool pieces the event logs from different systems together to create a model of your actual, as-is process, including the duration of each step.

Builds a visual representation of a business process: When you mine a business process, one output is a diagram of that process. This is called the discovered model. Models help to identify common bottlenecks, deviations, and other impediments to successful and efficient completions. You can visualize how long each step takes, where workflows stall and where the process goes after each step is completed. Perhaps there are unnecessary loops or paths that end before they should. You’ll be able to see those.

A view into workflows across systems, along with any process variants: A typical enterprise process has workflows, or activities that span more than one system. Here’s an example: a customer places an order for a large piece of equipment via paper form. An employee inputs the data into the CRM system. The system checks the inventory to determine shipping times and sends out confirmation to the customer via email—notice that we’re only a few steps into this process, and we’ve already passed through three systems. Process mining spans these systems to show you each step.

This example process sounds fairly straightforward, but process mining analysis can uncover hidden bottlenecks and variants that slow progress. In the example above, the employee has multiple forms to process, and may not get to all of them for days, leading to complaints. Also, depending on regional rules and regulations, the employee may have to follow a separate process path—another variant that takes time and increases the cost of the process.

Identify process inefficiencies and areas ripe for optimization: The delay in the example above, and the reasons for it, is just one illustration of a process inefficiency or bottleneck that could be discovered using process mining. Once discovered, this same process could be optimized using an automated solution like intelligent document processing (IDP).

Once you’ve visualized your process as it really works and identified the issues, you can build a target model to represent the ideal reference model for your process. You can upload the target model as a .BPMN file or create it directly in the software.

Comparing your target model against your discovered model (called conformance checking) helps you identify key variants and diagnose activity changes to optimize for the greatest efficiency. In highly regulated industries, for instance, comparing models is an easy way to identify and address process deviations as part of compliance.

7 misconceptions about how process mining works.

Now that you’re familiar with how process mining tools work, what don’t they do? Here are seven common misconceptions debunked.

Misconception #1: Process mining replaces human workers.

Because process mining analyzes data from event logs, it can replace the often-lengthy (and error-prone) portion of process improvement where a process manager or analyst has to interview stakeholders, shadow employees, and dig up data about time spent on each step. This software saves so much time—but does it mean the human worker becomes irrelevant?

The first thing to remember is that process mining is a tool, not an employee. It’s a set of capabilities that allow you to get faster, more accurate insights. Rather than replacing employees, this tool lets them spend less time tracking down process steps and more time focusing on making the process better.

Second, a process mining tool produces insights, but you still need a team who can interpret those insights. Even more importantly, you’ll need a plan for turning analysis into action. This will likely include ongoing design, automation, and optimization, along with coordinating change management, and applying those lessons learned to other processes. Ultimately, this investment will build the capabilities of your process managers and analysts.

[ We created a helpful primer on process mining. Get it now: Process Mining Guide ]

Misconception #2: A process can only be mined once.

Process mining provides opportunity for continuous optimization. It’s more than a single improvement project with a start and end date. You might think that you’re finished once you’ve mined a process and optimized workflows. But actually, you can and should re-run your analysis regularly.

Process mining can help you understand whether you’re reaching the goals you set for the new and improved process, as well as make it easier to identify additional areas ripe for optimization. You'll also catch any new issues—for example, did you create a loop unintentionally, or add an unexpected bottleneck? In this way, humans and this technology can work hand-in-hand to build an ongoing cycle of process improvement. Discover, optimize, and monitor, and then apply your learnings to even more processes across your organization.

Misconception #3: Process mining analysis is subjective.

A process mining tool replaces the subjective elements you find in traditional process optimization methods. Instead of interviewing stakeholders, tracing the steps, and tracking down the source of delays, your employees have an objective, data-driven way to view the process as it actually works, not how they think it works.

The input comes from your business systems, which keep track of the who, what, and when activities taking place in an event log. Using this information, the tool replicates your existing process so you can see exactly what’s happening. With AI-driven functionality and Appian Process Mining, you can use root cause analysis to understand the why behind process bottlenecks and inefficiencies, so you can build new workflows with confidence.

Misconception #4: You get deep insights into every process at once.

Process mining isn’t something you can use at a global level. Meaning, you can’t tell it to analyze every process in your company at the same time. It works best on one process at a time, analyzing and drawing out insights. Of course, you can mine multiple processes this way, and you should. But the tool isn’t designed to automatically generalize process issues simultaneously across your organization.

Here’s why this makes sense: A single process may have many subprocesses. In fact, at an enterprise level, most processes do. It’s much easier to analyze complicated processes on a smaller scale, and then apply those learnings to different areas.

Misconception #5: Insights and analysis act like a performance review.

Insights from process mining will not reveal who is doing a poor job at your company. Rather, you might find that you have a bottleneck in the shipping department, and you can dig into why activities are delayed.

Rather than revealing employee performance issues, process mining provides the opportunity to diagnose and fix inefficiencies, which helps workers do their jobs better. Maybe an underperforming process restricts resources, and employees have to wait days to execute requests. Maybe the approval process moves slowly. You can identify the real issues behind problems, so you can get them fixed.

Misconception #6: Process mining fixes all process deviations by itself.

Process mining is not a panacea for your process issues. Rather, it’s a discovery tool, like an X-ray or an ultrasound. By using it, you gain visibility and transparency into what’s going on.

Once you’ve gained insights, you can take the necessary steps to correct and improve process workflows. Automation tools like AI-powered IDP can download documents and extract information at a rapid pace, while robotic process automation (RPA) can take monotonous, time-intensive tasks off employees’ hands. A unified platform like Appian has workflow and process mining technologies built in (plus it’s low-code), so you can work quickly and easily, moving from insight to action and back again as you design, automate, and optimize your processes.

Misconception #7: Process mining technology is a passing fad.

Is this one of those tech trends that will fade away? We don’t think so. Polaris Market Research predicts that the global process mining software market will grow at a compound annual growth rate of 49% by 2030. This technology fills a gap in the market as businesses realize they need to add continuous optimization into their business process management strategy, especially if they want to keep up with or surpass their competitors. It’s about using the data you have to do that. While there are numerous ways to cut costs and improve performance, process mining is one of the first objective, data-backed strategies to help you understand what’s really going on.

Ready to transform your business with Appian Process Mining? Learn how easy it is to create a cycle of continuous improvement for a competitive advantage today.

[ Here’s a full breakdown on process mining: Process Mining Guide ]