Forum Discussion
Philosophical/preference question about planning and organization
Hi, I'm about to take a PL-300 course and I'm super excited. But until then, my Power BI service has started to look *very* messy and unorganized.
It started simple: I had a desktop report, I published it, then I did that a dozen more times
At this point everything was still very clear... I'd have one data model and one report file per report and everything was fine.
Then I started to use DataFlows. Wow are they cool. But now I'm using DataFlows. And this is causing a problem (in my brain) on how to organize the work. It's important to know that I work with Quality data. And so there are a ton of very repetative or similiar Quality Measures or Key Performance Indicators that I'm dealing with.
My old strategy was to have ONE desktop report that had about 8 KPIs in it. But a colleague suggested instead setting up 8 dataflows -- 1 for each measure -- and then trying to combine them later in a desktop report or something else. And I know I could use folders too but then I don't want to have to go DIGGING through so much to find stuff. I also like having everything in one page so I can look for the error /!\ symbols.
Whatever I do -- 8 is very managable. But we are soon expanding to about 30. And I don't know if grouping them or having 1 dataflow for each KPI is best. Or what the pros/cons of either would be.
Any ideas? How would you organize 30 KPIs?
Hi tnemeth ,
If you use a single dataflow per kpi
Pros: easy to manage and update individual KPIs; clear separation of data.
Cons: large number of KPIs can become overwhelming; difficult to get a holistic view.If you use grouped dataflows
Pros: fewer dataflows; easier to manage related KPIs together.
Cons: May require more complex data modeling; potential for large dataflows.You can choose based on your situation and usage preferences.
Also, you can find a white paper on dataflow at this link. This white paper describes dataflows in technical detail, and describes the capabilities and initiatives behind dataflow features and functionality.
White papers for Power BI - Power BI | Microsoft Learn
Best Regards,
Clara Gong
If there is any post helps, then please consider Accept it as the solution to help the other members find it more quickly.
2 Replies
Hi tnemeth ,
If you use a single dataflow per kpi
Pros: easy to manage and update individual KPIs; clear separation of data.
Cons: large number of KPIs can become overwhelming; difficult to get a holistic view.If you use grouped dataflows
Pros: fewer dataflows; easier to manage related KPIs together.
Cons: May require more complex data modeling; potential for large dataflows.You can choose based on your situation and usage preferences.
Also, you can find a white paper on dataflow at this link. This white paper describes dataflows in technical detail, and describes the capabilities and initiatives behind dataflow features and functionality.
White papers for Power BI - Power BI | Microsoft Learn
Best Regards,
Clara Gong
If there is any post helps, then please consider Accept it as the solution to help the other members find it more quickly.
Thank you very much: I was hoping for either a response like yours OR somebody saying "Oh no, you need to do it xyz way because that's the only good effective way."
It looks like I'll be going in the direction of *many* dataflows. One per KPI and then even some of those dataflows may have supporting dataflows... The reason for this is because I spent 3 hours last night trying to trace an error NOT present in the query editor but present when trying to refresh. But everything is a new build so I have no idea why or how that's possible.
Anyway I figure MANY queries are going to be much easier to identify, isolate, and correct issues. Which I'm bound to have.
Looking forward to my PL-300 course... haha.
