question

Oh boy... It might depend on the size of your organisation and the resources they can throw at their BI team and solution. If you're a small team in a mid-sized organisation, you might have to prioritise what you can learn (and therefore deliver). It's an incredibly broad and fluid field these days. I don't know how anyone could stay abreast of all of it, even if you just confined yourself to the Microsoft stack. So rather than try and answer your question directly, I might just ask some questions to see if I can provoke some ideas for further enquiry. Do you have a good understanding of star schemas and how to transform data from snowflake (best case scenario) schemas and other structured data sources (for example XML, JSON, SOAP, delimited text files)? Do you understand why and how to do incremental loads? Can you trace your atomic facts back to their source rows if you have to? Do you know what a slow changing dimension is? Have you ever needed to implement slow changing facts? Do you know your way around BIML? What about Powershell? Do you have automated testing set up (such as nUnit - that's not an endorsement BTW) to ensure your transformed data reconciles with the source? Moving from the ETL side to analysis, are you familiar with the R language? When would you favour Multidimensional SSAS over Tabular? Where are you at with MDX and DAX? DMX anyone? What do you know about Power BI (and PowerQuery's "M" language)? Do you use version control (such as TFS)? Are you across the data and information governance issues that your organisation will need to embrace to successfully implement a BI solution. In this sphere you might look at Data Quality Services? Is your organisation integrating it's BI solution in SharePoint? There's a whole can of worms right there. I could go on, but suffice it to say that I suspect that where small teams DO implement successful BI solutions, they probably do so by starting small and doing the fundamentals well rather than jumping at all the shiny new toys that get pitched around. It's hard to know which technologies will be the best ones to learn, and it's even harder to say "No" when your CEO says "I want my dashboard to look just like [insert massively impressive demo here]". Hope this helps :-)