• Link Orders to Regional Manager using the common Region column.

• Link Orders to Returns using the `Order ID` column.

Step 4: Convert Tableau LOD Expressions to Power BI DAX

This is where the transition gets interesting. In the original Tableau dashboard, a cohort retention metric relies on a Level of Detail expression to isolate the exact date a customer made their very first purchase:

{ FIXED [Customer ID] : MIN([Order Date]) }

Power BI does not use a FIXED keyword. Instead, Claude translates this calculation into an explicit DAX calculated column using the CALCULATE and ALLEXCEPT modifiers.

1. Go to the Data View or Report View in Power BI.

2. Select the Orders table from your fields pane, right-click, and choose New Column.

3. Paste the DAX calculation script provided by Claude:

First Purchase Date =
CALCULATE(
    MIN(Orders[Order Date]),
    ALLEXCEPT(Orders, Orders[Customer ID])
)

This expression instructs Power BI to evaluate the minimum order date while clearing all filters on the Orders table except for the Customer ID. This mirrors the Tableau behavior perfectly.

Step 5: Build the Customer Retention Visualization

With our data model ready and our DAX column calculated, we can now configure the final dashboard presentation. We will create a stacked column visual to review how historical customer cohorts continue to purchase goods across subsequent operational years.

1. In the **Visualizations** pane, select the Stacked Column Chart icon.

2. Drag the standard Order Date from your fields pane over to the X-axis well. (Right-click it and ensure it is displaying the date hierarchy so you can isolate just the Year value).

3. Create a clean explicit measure for your metric tracking (e.g., Total Sales = SUM(Orders[Sales])) and drag it to the Y-axis well.

4. Finally, grab your newly calculated First Purchase Date column and slide it directly into the Legend well. Make sure this field is filtered to display the Year as well.

If the output initially looks like a grouped side-by-side bar layout, double-check that your visualization type is specifically switched to a Stacked Column Chart to properly layer your consumer retention categories.