An analyst's main goal is to spot trends, performance and communicate insights faster from provided data. This makes them seek a software that supports data cleaning, create impactful visuals and present the data in a way that actually tells a story and understand the data better. Hence, Power Bi.
How do analysts use Power Bi?
1. Get the data
First, the analyst sources for the data. They may get the data from the company’s database, it may be in Excel format or CSV or it may come from POS machines or even surveys.
2. Power Query in Power Bi
They then transform the data in Power Query where they get to clean it. Clean it? Is data dirty? Yes! Raw data collected from a source is very messy. Data can be messy in a number of ways, these include:
- Missing records: these may be resolved by filling in the values with unknown for text data types and null for numeric data types.
- Duplicate values: Analysts remove the duplicate rows to ensure accuracy in the data.
- Inconsistent data for example, you may find a column such as city with values such as NRB and Nairobi. These values are the same but to the software, they appear as different cities. The analyst will standardize the data by restructuring all formats to be consistent.
- Wide tables: The analysts Unpivot the columns to normalize the data.
- Wrong data types: The data types of the columns are adjusted in Transform tab.
- Use DAX to do calculations. In order to create insightful reports, I can use DAX functions to perform dynamic calculation on the data. FOR EXAMPLE SUM() which adds up all the values in a row. Create a new measure.
Total sales= Sum(Sales[Amount])
What if you want to get the expected results, multiple columns are required? Then you would use SUMX()
Total Revenue = SUMX(Sales,Sales[Quantity]*Sales[Price])
What about calculating KPIs? Then I would use Calculate()which allows me to evaluate an expression with complex filters.
Total Sweater sales 2024= CALCULATE(SUM(Sales[Amount]),
Sales[Product] = "Sweater", YEAR(Sales[Date]) = 2024)
4. Create Visuals
Analyst then build visualizations from the functions conducted. There are so many types of visualizations in Power Bi. You can either use reports or dashboards.
Reports are used for detailed analysis and can have multiple pages and visuals while Dashboards are most used for summaries therefore only contain the most important insights.
For the functions performed above, the most preferred would be a Card. It is used to visualize a single important values.
Other visuals that are very popular include:
- Bar/Column charts: These are used for comparing values in different categories.
- Pie/Donut charts: Used for showing the percentage of a value to a whole.
- Slicers: For filtering the visuals on the dashboard.
- Map visuals: Ideal for geographic analytics either filled or bubble map.
5. Present Findings
An analyst then publishes their findings for executives. These can be done using Power Bi Service where a company can host their own servers securely. There is also a mobile App that allows one to view their reports on the go.
Summary
Power Bi has proven to be a very necessary tool for every data specialist. It can assist in converting chaotic data into very compelling insights. In order to story tell using data, mastering Power Bi is nonnegotiable as a data analyst.
Top comments (0)