How to import a CSV file into SAP Analytics Cloud Modeler

Importing a CSV file into SAP Analytics Cloud is one of the first operations every analyst should master. Done right, it takes five minutes. Done wrong, it produces models with broken aggregations and unreadable dimensions. This guide covers the full procedure.

Prepare the file before importing

A successful import starts with a clean file. Before opening SAC, check that your CSV meets these conditions: a single header row, descriptive columns (dimensions) on the left, numeric columns (measures) on the right, no merged cells and no blank rows in between.

The encoding must be UTF-8 to preserve special characters. A file exported from Excel as "CSV Windows" often uses Latin-1 encoding that turns "Région" into "Région" on import. Re-export as "CSV UTF-8" if needed.

Open the Modeler and start the import

In SAP Analytics Cloud, open the main menu then Modeler > New Model > Import data from a file. Select your CSV file. SAC shows a preview of the first rows and tries to guess each column's type.

This is the most important step: never blindly trust automatic detection. SAC regularly gets two cases wrong — leading-zero codes (which it classifies as a measure) and date columns (which it sometimes leaves as text).

Map the dimensions

A dimension is an analysis axis: a region, a product, a channel, a period. Drag each descriptive column to the Dimensions area. For the time column, explicitly set the Date type — without it, you lose all year-to-date and prior-year comparison functions.

Beware of identifiers: a product code like "00712" must be typed as a Dimension (String), never as a measure. If you leave it numeric, SAC strips the leading zero and tries to add up codes, which makes no sense.

Map the measures and choose the aggregation

A measure is an aggregatable value: revenue, quantity, rate. Drag the numeric columns to Measures. For each, check the default aggregation. SAC suggests SUM, but that's only correct for amounts and volumes.

For a conversion rate, a percentage margin or a score, choose AVERAGE: adding percentages produces nonsense (a 2.8% rate over a hundred rows would become 280%). For a stock or a balance, use LAST to take the end-of-period value rather than their sum.

Validate the types and save

Before saving, go through each column and confirm its type: Date for the period, Decimal or Integer for measures, String for codes and labels. Once the model is validated, click Save and give it a clear name.

Build your first Story

The model is ready. Go to Files > New Story and select it. Add a first chart: a bar chart to compare values by dimension, a line to track a trend over time, a donut to visualize a breakdown. Add filters on your main dimensions to make the Story interactive.

Save time with a ready-to-use template

If you're new to SAC or want a credible starting point, starting from an already-structured template skips this whole modeling phase. The SAC Templates Hub catalog offers 64 models across 16 industries, each with its KPIs, dimensions and aggregations already correctly configured. You download the file, import it following the steps above, and get an operational Story in minutes.