Importing a multi-sheet Excel workbook into SAP Analytics Cloud

Importing a multi-sheet Excel workbook into SAP Analytics Cloud trips people up not because the upload is hard, but because a workbook is structured for humans and a SAC model is structured for a machine. Get the preparation right — one clean fact sheet, explicit dimensions, the correct aggregation on every measure — and the import is a five-minute job. Get it wrong and you spend an afternoon chasing wrong totals. This guide walks the full path, from preparing the file to validating the result.

First, understand what SAC wants

A SAC model is built from dimensions (the things you slice by — Entity, Account, Period, Product) and measures (the numbers — revenue, headcount, days). When you import an Excel file, SAC needs to map each column to either a dimension or a measure, and it needs each row to be a single fact. A typical analyst workbook, with merged cells, sub-totals, and a different layout per sheet, fights that expectation. The work is in reshaping the workbook so SAC can read it cleanly.

Step 1 — Prepare the workbook

Decide which sheet is your fact data — the granular transactional rows — and treat the others as supporting data (master data, mappings, a KPI summary). On the fact sheet:

• One header row at the top, with a clear, unique name per column. No merged cells, no sub-headers.
• One fact per row. Remove sub-totals and grand totals — SAC will compute those itself from the aggregation. Leaving them in double-counts your data.
• A column for every dimension. If a value is implied by the sheet name (say the sheet is "2026" or "France"), add it as an explicit column instead, so the dimension travels with the row.
• Clean, consistent values in key columns — the same entity spelled the same way everywhere, dates in a single format.

If the data you need is spread across several sheets with the same structure (one per month, one per region), stack them into a single fact sheet with an extra column identifying the source (Month, Region). One tall table imports far more reliably than twelve wide ones.

Step 2 — Import into the Modeler

In SAC, open the Modeler and create a new model (or open an existing one) and choose to import data from a file. Upload the workbook and, when prompted, select the sheet that holds your fact data. SAP’s own import documentation on help.sap.com walks through the dialog box by box. SAC reads the header row and presents each column for mapping. This is the moment that decides whether your model is correct.

Step 3 — Map columns to dimensions and measures

Go column by column. Assign each descriptive column to a dimension (Account, Entity, Period, Product, and so on) and each numeric column to a measure. Map your time column to the model's Date/Period dimension explicitly so SAC understands the time hierarchy — this is what lets you roll figures up by month, quarter and year later. Validate the data types SAC infers: a code that looks numeric (a cost-centre ID) should usually be a dimension, not a measure, or it will be summed nonsensically.

Step 4 — Set the aggregation on every measure — this is the one that bites

For each measure, set how it aggregates across the time hierarchy. Additive figures — revenue, units, emissions — use SUM. But balances and rates do not: a headcount or a capital balance should aggregate with LAST (the closing value), and a ratio should never be summed or averaged across rows at all — it must be a calculated measure. This single setting is the most common cause of "the import worked but the year total is four times too big." If you are unsure which to pick, read how to choose the right aggregation in SAC before you save — it is the difference between a model that reconciles and one that quietly lies.

Step 5 — Validate, then build

Before trusting the model, sanity-check it: does the grand total match the source workbook's total (after you removed its sub-totals)? Spot-check one entity and one period against the original. Confirm the time dimension rolls up correctly month to quarter to year. Only then build your Story on top. A five-minute validation here saves you from publishing a dashboard that is confidently wrong.

Common pitfalls

• Leaving sub-totals in the data — they get imported as facts and double-count. Strip them.
• Implicit dimensions in sheet names — make every dimension an explicit column.
• Default SUM on balances and rates — set LAST for balances; make ratios calculated.
• Inconsistent member names — "FR", "France" and "FRA" become three different members.
• Numeric codes as measures — IDs belong on a dimension.

Skip the modeling step

Our ready-made templates download as .xlsx already structured into Data, KPI Summary and model-definition sheets, with dimensions and aggregations already correct — for example the month-end close template is a clean multi-sheet workbook ready to map. They import through exactly this procedure, so you skip the reshaping entirely. If you start from a CSV instead, see importing a CSV into the SAC Modeler.