A 90-day plan to learn Power BI from scratch

Power BI is easy to start and surprisingly hard to do well. The reason most learners stall is not the tool, it is the order they learn it in. Here is the sequence I recommend, broken into three months.

Month 1: get the data into shape

Spend the first four weeks inside Power Query, not on charts. Learn to connect to Excel and CSV files, remove junk rows, fix data types, split and merge columns, and append files from a folder. A report built on clean data takes minutes; a report built on messy data takes forever and breaks on refresh.

• Week 1: import, transform and load a single sheet.
• Week 2: merge two tables and understand the join types.
• Week 3: append twelve monthly files into one table automatically.
• Week 4: document your steps and refresh the report to prove it survives new data.

Month 2: model before you visualise

This is where good analysts separate from beginners. Learn the star schema: one fact table of events, surrounded by small dimension tables for customers, products and dates. Build proper relationships, add a date table, and only then write your first DAX measures. Start with SUM, DIVIDE and CALCULATE, and make sure you can explain filter context in your own words.

Month 3: design for decisions

Now the visuals. Pick one business question per page. Lead with the headline number, show the trend, then the breakdown. Remove every element that does not help someone decide something. Finish by publishing to the Power BI service, setting up a scheduled refresh and sharing the report with a real person for feedback.

What to have at the end

One portfolio project you can walk through in ten minutes: the business question, the messy source, your data model diagram, three measures you are proud of and the final report. That single project will do more for your job search than ten certificates.