Guides

How to Connect Excel and CSV to Power BI Desktop

By Ravindra Bagale, Power BI & Data Analytics trainer · LinkedIn · Published: 30 September 2026 · Last updated: 30 September 2026

Connect Excel or CSV in Power BI Desktop with Get data, preview the table in Navigator or Power Query, fix headers and types, then Load (or Close & Apply) so the data sits in your model.

File → Get data → Preview → (optional) Power Query fixes → Load

What do I need before this guide?

• Power BI Desktop installed (install guide).
• An .xlsx or .csv file on your machine (create a tiny one if needed).
• Course pointer: Get data window.

Before and after (look at the tables first)

Connect Excel/CSV, clean in Power Query if needed, then Close & Apply. Here a Year = 2025 step leaves 140.

Why start with Excel and CSV?

Because they teach the full connector habit without a database password.

Why you care:

1. Real offices still email Excel every week.
2. Ops tools often dump CSV.
3. You learn Get data, preview, Load vs Transform, and data types early — the silent killer of wrong charts.

Once this muscle is built, SQL and Folders feel like the same movie with a new scene.

Tiny example we will aim for:

1. Mumbai Amount 100, Pune Amount 40.
2. After Load, a Card shows 140.
3. A City slicer set to Mumbai shows 100.

Excel workbook — step by step

What this is: pulling sheets or named tables from an .xlsx file into Power BI.

Why you care: Excel is the most common first source in Indian offices and training labs.

Vertical steps:

1. Open Desktop → blank report (or your practice .pbix).
2. Home → Get data → Excel workbook.
3. Browse to the file → Open.
4. Navigator lists sheets and tables.
5. Tick the object that holds the rectangular data (prefer a named Table).
6. Preview on the right — check column names.
7. Click Transform Data if the first rows are titles/notes; otherwise Load.

What you see on screen:

1. A Navigator window with ticks on the left and a preview grid on the right.
2. After Load, Fields shows your table name.

Sheet vs Excel Table (why it matters)

1. A sheet may include logos, titles and blank rows above the real grid.
2. A named Excel Table is a structured block Power BI loves.
3. If you control the workbook, Format as Table in Excel first — future you will thank you.

Tiny Excel example (FreshBasket, fictional)

File freshbasket_orders.xlsx:

1. Columns: OrderDate, City, Category, Amount.
2. Two data rows: Mumbai / Snacks / 100, Pune / Dairy / 40.
3. Named table Orders.
4. In Navigator, tick Orders only — skip the “Cover” sheet with the title art.

What happens after Load:

1. Data view shows two rows.
2. Card on Amount shows 140.
3. Column chart by City shows Mumbai 100 and Pune 40.

CSV files — step by step

What this is: importing a plain text table where commas (or another delimiter) separate columns.

Why you care: many tools export CSV, not Excel.

Vertical steps:

1. Home → Get data → Text/CSV.
2. Select the .csv file.
3. In the dialog, inspect:
4. Delimiter (comma, semicolon, tab, …) — the character that splits columns.
5. Encoding (UTF-8 is common; try another if Marathi/Hindi text looks broken).
6. Data Type Detection (you can be strict and set types yourself in Power Query).
7. Click Transform Data for real work, or Load for a perfect file.
8. In Power Query: promote headers if needed, set types, remove junk columns.
9. Home → Close & Apply.

What you see on screen:

1. A preview of columns split correctly (City in its own column, Amount in its own column).
2. If wrong: one sad column with commas still inside the text.

When CSV becomes one sad column

1. The file used ; but Power BI expected , (or the reverse).
2. Fix delimiter in the CSV dialog, or Split Column by delimiter in Power Query.
3. Re-check the preview before Close & Apply.

Tiny CSV sample:

OrderDate,City,Category,Amount
2025-09-01,Mumbai,Snacks,100
2025-09-01,Pune,Dairy,40

After a clean Load, Card = 140.

Load vs Transform — the decision

Use this vertical rule:

1. Preview looks perfect (headers + types) → Load.
2. Anything smells wrong (merged title row, wrong types, extra columns) → Transform Data.
3. When unsure → Transform. A two-minute fix beats a broken model all week.

What Transform is: opening Power Query (the cleaning window) before the table enters the model.

Why you care: wrong types make “100” act like text, so totals fail.

Power Query details expand in Power Query vs DAX and the course Power Query chapter.

After the data lands

Vertical steps:

1. Open Data view — spot-check rows (Mumbai 100, Pune 40).
2. Open Model view — one table is fine for day one.
3. Build a Card on Amount and a slicer on City.
4. Click Mumbai on the slicer — Card should show 100.
5. Save the .pbix.
6. If the Excel path moves later: File → Options and settings → Data source settings → change source.

What you see on screen: Card 140 with no slicer; Card 100 with Mumbai selected.

Common errors and fixes

Worked Excel story (fictional FreshBasket)

File layout:

1. Sheet Cover — logo and “Confidential training sample”.
2. Sheet Orders — messy title in row 1, headers in row 2.
3. Named table OrdersTable on a clean rectangular range (best) with Mumbai 100 and Pune 40.

What you do:

1. Get data → Excel → tick OrdersTable only.
2. If you must use the messy sheet: Transform → Remove top rows → Use first row as headers → set types.
3. Close & Apply.
4. Never load Cover.

What you see: Fields shows OrdersTable; Card shows 140.

Worked CSV story

orders_2025-09.csv:

OrderDate,City,Category,Amount
2025-09-01,Mumbai,Snacks,100
2025-09-01,Pune,Dairy,40

1. Text/CSV → check comma delimiter.
2. Transform → Date type on OrderDate, Decimal on Amount.
3. Close & Apply.
4. Card + City slicer proof (Mumbai → 100).

Folder of CSVs (preview only)

When monthly files share the same shape, the course Folder connector combines them. Master one file first; Folder is the encore.