09/15/2023
I heard the terms CONVERGENT vs. DIVERGENT thinking pop up in a video that was shared with me by Alex Hormozi (bring out the pitchforks 🔱🔥🪧).
It's a common distinction that was coined in the 1950s by a psychologist. And it reminded me of why Data Analytics is misused and hard to get right. Especially in technical organizations like many Manufacturers.
Many engineering problems in manufacturing can be solved with convergent thinking (a singular correct solution to a problem). You know your inputs and your desired outputs. You can tweak some variables to get the desired output.
And that certainly isn't incorrect in any way. It just can be hard to get out of that mindset when you need to.
With data you need convergent thinking to build those dashboards for monitoring things. Things like OTD, revenue, OEE, and other KPIs to keep an eye on.
But to be "data driven" and give high impact analytics, you need to think divergently (coming up with many potentially correct solutions that can be chosen from).
And it's actually a cycle.
1. You use convergent thinking to define a problem. e.g. our OEE is too low
2. You use divergent thinking to get data around Availability, Performance, and Quality. This data gives potential solutions to fix that bad OEE metric.
3. You use convergent thinking to choose the solution that the data suggests gives the best and biggest chance of success.
4. You implement the solution, and use divergent thinking to gather data on all the potential downstream impacts of the change.
5. If it didn't work or something negative and unexpected happened (e.g. Availability improved but Quality dropped), you use convergent thinking to narrow in on the problem again. AKA start back at 1.
If you get stuck at the Convergent thinking stage, you end up asking "so what" when talking data.