BI analytics: where to start?

Question

IT education

I have been working as an internet marketer for over 3 years. It took me a long time to make the decision to switch to BI analytics. But, as it seems to me (and most likely it is), I see an incomplete picture and cannot imagine the full stack of necessary tools to start self-study, so I began to consider paid courses from well-known training companies. Are there people here who entered the profession through such courses? Could you share your experience, are they worth it, or is it a waste of time and money, and would it be more correct to continue self-study?

Answer 1

The courses are all different, you need to look at yourself to make it clear and interesting.
I will not recommend courses, I will only describe technologies in order of importance, as well as BI tools that you should start working with.
1 and most important - SQL. Without it, nowhere. There are many specific DBMSs with their own dialects, but it is better to start with some mature DBMS that has a developed dialect and good compatibility with the SQL standard, such as postgresql. But if a database is already used at work, it will be the answer. In analytics, other databases like Clickhouse are more popular, but they are limited in terms of language capabilities and it is better to understand than not to understand. With large amounts of data, of course, you will have to work with databases tailored for analytics.
2 and no less important - statistics. At a minimum, you need to understand what A / B testing is, how to conduct it, how to correctly put forward hypotheses and test them. Significance
3 - what you can do without, but what you still have to learn if you don’t want to know more and move further - python (+ pandas + jupyter and mb some other libraries and frameworks)
4 - something that is not something as complicated as python or matstat, but will be needed in the work - Bi platforms. Mainstream are tableau, power bi and qlik. Tableau is the most versatile option with good rates and the ability to fully explore the tool for free. With him, I advise you to start. Power bi is a close competitor from Microsoft (which limits the technology stack a little, but in general everything is ok). C qlik did not work, but it feels like a very old platform used by companies that are not the most "IT" and flexible in terms of technology. It is also very powerful, but it is rather an outsider.
Also noteworthy are the open source (and free) Apache superset, metabase and redash. If the company does not have a lot of data and it is not too complex, I can advise you to start implementing BI with metabase - a convenient system that is being actively developed. With a well-designed database, it allows you to watch many things without SQL at all, as well as pump SQL well on more complex tasks. Of the minuses - not a very large number of visualizations (however, all the most important things are there) and a not very clear system of privileges and access rights. However, it is already very good that the open-source BI has such a developed system of access rights. Redash is like a meta. The superset will surpass them in the future, but why is it needed if you can’t build a regular chart on which there is nothing on the X-axis except dates .. however, mb this problem has already been completed
There is also Google Data Studio, which has excellent integration with Google services, but it seemed to be damp, however, this is a Google product and this is a sign of quality.

Answer 2

Get started with Microsoft Power BI (it's free for private users). A lot of courses and textbooks (at least on Habré)