Reporting on ROAS in GA4: importing your cost data

Beyond the tracking code on your site or app, GA4 lets you set up scheduled imports for all kinds of data: user IDs, item data, offline events and cost data.

The last one is what matters here, because without cost you can't calculate ROAS.

First, return calculations that hold up. Know exactly what you spend on advertising and you can work out what it returns, and decide where budget goes on evidence rather than instinct.

Second, deeper campaign analysis. With cost data alongside, you can track performance down to campaign, ad set and even individual ad level, and see what actually drives revenue.

• Google Sheets
• Amazon Redshift
• Amazon S3
• BigQuery
• Google Cloud Storage
• HTTPS
• MySQL
• PostgreSQL
• Snowflake

• Go to Admin, middle column, data import.
• Choose create data source, then cost data.
• Pick your import source: manual CSV or SFTP.
• Download the template from the Google help article and stick to its formatting.
• Update your UTMs so they match, and add the UTM ID. Without that link GA4 can't connect the ad click to the session.
• Export your data from Microsoft Ads, Meta Ads or wherever you advertise.
• Map the analytics fields to the import fields.

Once you've created the data source, the field mappings can't be changed. Check them carefully before you save.

GA4 accepts one import source. If you want data from several platforms, combine them into a single sheet. Keep that file somewhere sensible, because you'll be updating it regularly.

The import itself takes a while. Come back to it later and verify every field.