Guidelines for understanding a model created by someone else

[BlueNote]

Hi Everyone,

I am looking for any sort of advice or guidlines on taking over Power BI solutions created by other people.

In some cases I am taking over solutions that have bad practices embedded in them. As I start working with these models the bad practices often pop up and I'd rather have a process to discover them in advance if possible.

Is anyone aware of a guide or article or something that explains a good method for mapping out a Power BI solution that has been created by someone else? Something that focuses on understanding what is happening and documenting it. 

I did my best with a google search and didn't find much.

[jdbuchanan71]

My first thought is, before you try to make any changes, save a backup of the reports as they are right now and never change the backups.  That way, if you change something that breaks some reports or visuals, you can always go to the backup to see how it was in the original and you can find what you need to fix.

There is a 'Best Practices Analyzer' in Tabular Editor that can be handy.  At the end of the article there is a link to a GitHub page where they have a set of rules you can import and run against your model.

https://docs.tabulareditor.com/te2/Best-Practice-Analyzer.html

The PowerBI Field Finder can help identify items in the model that are used in reports.

https://powerbi.tips/2020/01/power-bi-field-finder/

I would also hook the Vertipaq Analyzer to the model so you can see all the of the tables / fields / measures and see what is taking up the most space in the model.

https://www.sqlbi.com/tools/vertipaq-analyzer/

[Nathaniel_C]

Hi @BlueNote ,

You are dealing with an issue that can be a big problem, especially due to the lack of documentation that accompanies a report.  I add various hidden pages with notes for my projects.

At my work, if I take over a project, if it is working, I try to make as few changes as possible, and let my manager know that it is a poorly created report.  The most important failing is typically the data model, although I have also had to deal with "faked in" graphs etc.

You will have to weigh the time it would take to creating a new project, or working with the old one.  In some cases, I have worked minimally on the project, but find over time, as change requests come in that I have a better understanding of the data and process.

They are a ticking bomb, though. 

Let's ask some others.  @jdbuchanan71 , @ImkeF , @KenPuls @MattAllington 

Let me know if you have any questions.

If this solves your issues, please mark it as the solution, so that others can find it easily. Kudos 👍are nice too.
Nathaniel