The column names below are examples. Marketplace report formats change, so map them to the files you actually export.

1. The tables

Load each report as its own table (CSV exports work fine) and clean the column names in Power Query.

Tips for the load step:

• Use the same SKU text in every file (trim spaces, fix case). Most broken dashboards come from mismatched SKUs.
• Make sure Date is a real date type in every table.
• Create DimDate with CALENDARAUTO() or a date table from Power Query, and mark it as the date table.

2. The relationships

Use a star schema: each dimension connects to each fact table with a one-to-many relationship, filtering in one direction (dimension to fact).

• DimDate[Date] to FactSales[Date], FactAds[Date], FactInventory[Date]
• DimProduct[SKU] to FactSales[SKU], FactAds[SKU], FactInventory[SKU]
• DimMarketplace[Marketplace] to the Marketplace column of each fact table

Do not connect fact tables to each other. Slicers on the dimensions then filter all of them at once.

3. Core measures

Total Sales = SUM ( FactSales[Sales] )

Units = SUM ( FactSales[Units] )

Ad Spend = SUM ( FactAds[Spend] )

Ad Sales = SUM ( FactAds[AdSales] )

ACOS = DIVIDE ( [Ad Spend], [Ad Sales] )

ROAS = DIVIDE ( [Ad Sales], [Ad Spend] )

TACoS = DIVIDE ( [Ad Spend], [Total Sales] )

What they tell you:

• ACOS is ad spend as a share of ad-attributed sales. Lower is better.
• ROAS is the same relationship flipped: sales per dollar of ad spend.
• TACoS compares ad spend to total sales, so it shows whether advertising is growing faster than the business.

4. Trend measures

Sales Prior Month =
CALCULATE ( [Total Sales], DATEADD ( DimDate[Date], -1, MONTH ) )

Sales MoM % =
DIVIDE ( [Total Sales] - [Sales Prior Month], [Sales Prior Month] )

Put DimDate[Month] on a line chart axis and these two measures will show growth and slowdowns without any manual comparison.

5. Inventory measures

Inventory is a snapshot, so you should not add it up across days. Take the last value in the selected period instead:

Ending Inventory =
CALCULATE ( SUM ( FactInventory[EndingUnits] ), LASTDATE ( DimDate[Date] ) )

Daily Units =
DIVIDE ( [Units], DISTINCTCOUNT ( FactSales[Date] ) )

Days of Cover =
DIVIDE ( [Ending Inventory], [Daily Units] )

Sell-Through % =
DIVIDE ( [Units], [Units] + [Ending Inventory] )

Days of Cover estimates how many days the current stock will last at the recent selling pace. Sell-Through % shows how much of the available stock actually sold in the period.

6. A simple report layout

• Top row: cards for Total Sales, Units, ACOS, ROAS, Days of Cover
• Middle: line chart of Total Sales and Ad Spend by month, with a Marketplace slicer
• Bottom: table by SKU showing Total Sales, ACOS and Days of Cover, sorted by Total Sales

Add conditional formatting to the SKU table, for example red when Days of Cover is below your reorder lead time.

7. Check your numbers

Before sharing the report:

1. Compare Total Sales for one month with the marketplace's own sales report.
2. Compare Ad Spend with the advertising console.
3. Check for SKUs that appear in the facts but not in DimProduct, since they will show as blank.

If the totals do not match, the cause is usually a date or time zone difference, or a missing file for part of the month.

8. Refreshing the data

For a first version, replace the CSV files in a shared folder and use the Folder connector so Power BI picks up new files. When you outgrow manual exports, move to scheduled refresh using the marketplace APIs or a data warehouse.

We build marketplace dashboards like this at Perfality. For a non-technical overview aimed at store owners, see our article Stop Guessing, Start Growing: Power BI for E-Commerce Success.