Process Mining vs. Process Discovery vs. Task Mining vs. Task Capture?

Automation and process analytics companies are great at naming new technologies. The hard part is to understand what those actually are and how those differ from each other. After that one can start to think if that specific technology helps with the problems in mind. In this post, I’m trying to clarify this space.

Let’s start with the good news: there are lots of excellent tools to help you. Many of those also end with a word mining - which has nothing to do with traditional mining, just like RPA has nothing to do with mechanical robots. (Disclaimer, process mining can naturally be used in traditional mining, and RPA be combined with mechanical robots, e.g. UiPath has a funny demo on that on YouTube:

Ok, back to the topic. Below is a summary of the terminology. The hardest term to describe is a process discovery since it has so many different meanings: e.g. BluePrism Process Discovery is about an “online survey that assesses and ranks your processes in term of most automation-ready to least” whereas Kryon Process Discovery is about automated process discovery where a “robot” identifies work processes and visually maps and evaluates those to be exported for their RPA Studio product. Automation Anywhere Discovery Bot has a similar target. Based on these findings, I would summarize that process discovery is basically task mining for RPA - similar thoughts have been in the Process Mining groups on LinkedIn.

Process mining

  • “Analysis of business processes based on event logs” -Wikipedia

  • Aimed for end-to-end process analysis, design, and enactment: trigger further operations

Task mining / process discovery

  • “Discovery, monitoring, and analysis of user interaction data on a desktop.” -myInvenio

  • RPA vendors aim using on speeding & scaling automation

  • Process mining vendors aim using on filling the gaps in system logs with work that happen on user desktop

Task capture / automatic PDD

  • “Document work just by doing it” -UiPath

  • Aimed for automated process documentation for as-is automation

As we quickly find out from the table, those different technologies are addressing very different problems.

The question is not about which technology to use, but rather what are you trying to achieve?

Process mining brings visibility into processes within enterprise systems, like ERP and CRM, which are typically already automated and have huge transaction volumes with high business criticality. Automated process discovery is about finding parts of process paths to be automated with RPA. Task capture is about documenting what the RPA bots are designed to be doing.

How does fit into these categories?

We at are close to task mining by definition, but with a different focus: identifying manual work across and around all systems regardless of the process. I would like to call it Work and task mining: spots all manual knowledge work

Workfellow spots all manual knowledge work

Objective identification and classification of manual work is essential to understand the best approaches to get rid of manual work and friction at scale.

Why does Work Mining matter?

Half of the knowledge work activities are solvable already with the existing solutions. I think it is safe to say that there is no magic wand that would make all these manual activities vanish overnight. Instead, we should be converting and making sure that as much as possible of the manual work is converted into automation platforms and core systems. Typical solutions that are identified with work mining are (not in a specific order):

  1. Correct settings and automation within existing systems

  2. Training, tips & tricks to make most out the current systems

  3. Scripts & low-code: small solutions touching and distributed to many

  4. Integrations and APIs

  5. RPA front office for attended automation

  6. RPA back office for unattended automation

  7. Lean and process/task development and optimization

  8. New SaaS software for specific problems

Data driven opportunities for Enterprise automation development

Thank you for your interest! Please let me know your thoughts, e.g. using LinkedIn:

I’m happy to network with you!



Follow us for the latest insights:

©2020 by