Process mining data requirements
What data is needed for process mining? 🤔
I hear this question a lot, so have put together a one-pager to help answer it.
Why is this needed? The way the story usually goes is that a business user at my client is interested in process mining, wants to get started, and now wants to know what data they need to request.
… but it’s not quite that simple.
There’s no pro-forma list of requirements that I can fire off to their data admin team.
– What’s the process?
– What are the questions about it?
– What data is even available?!
‘Nick, you’ve done this before. Just tell us!’ … but when you’re unquestioning of what data is available and how to get it, you’re asking for trouble. Never start a project without proper scoping first.
I’ve heard of horror-stories of getting a process mining use-case all lined up… but the data took months to arrive and then even longer to clean and transform. You *must* avoid this! 😲
– Go with the grain 🌾
– Pick low hanging fruit 🍎
– Take easy wins 🏅
…. pick your metaphor, but you have to understand the data landscape for each analysis to achieve rapid time-to-value with process mining. (Which is the goal / requirement!)
Stakeholders want answers asap:
– Address their critical business questions? 🤑
– What data is easily accessible to answer these questions? 💽
– Is other data available that answers questions they didn’t ask but should have?! 🤭
… then pack your event log and subsequent analyses with value and take the praise of your stakeholders for the ju-jitsu moves you made with the limited and messy data set you were given. 🥋
Know and spot data challenges so you can guarantee value with your analysis, and then over-deliver. ⚡
Leave a Reply
Want to join the discussion?Feel free to contribute!