Hey @paulfink ,
that's a difficult question as it depends in the scenario and the company.
When we have someone applying in my company, I am looking for the following things:
- First of all does the applicant understand and try to understand the data. Really make sure you know what you are talking about, as that is what will be your role later. Ask questions when something is not clear also during the process of creating.
- Create a proper data model in (best case) a star schema. Proper modeling is key to good analysis. Create fact and dimensional tables and create proper relationships.
- Work clean, for example hide ID columns when they are not needed, remove columns that are not needed (remove description columns in multiple languages when you analyze in English only)
- Try to figure out what the user of the report would be interested in. Put yourself in his shoes, what does he want to know, where is he usually struggling. Which insights are difficult to get with a normal data structure.
- Combine contexts. For example, show the history of a measure over time as tooltip when you just take a look at a specific moment in you analysis.
- Usually there are traps in tasks like that. Try to solve them and don't be shy to discuss them with your imaginary client. This will happen later all the time, be proactive, tell them the options and show you can communicate like a pro.
At the moment what I can think about, this seems to be the most important for me.
If I answered your question I would be happy if you could mark my post as a solution ✔️ and give it a thumbs up 👍
Best regards
Denis
