Adobe Analytics Classifications Importing metadata for collected dimensions

This page includes an overview of Adobe Analytics classifications, best practices for classification uploads, and a sneak peek into classification improvements on the horizon.

What are Classifications?

Classifications (formerly SAINT classifications) are created by grouping (classifying) granular data from a source dimension, such as any eVar, prop, marketing channel, etc. For example, you might want to analyze campaign codes by campaign group, or product IDs by product type. Below is a quick overview of classifications to get you started!

Automating Classifications

Classification Rule Builder

Do classifications seem like a lot of work? Don't worry, you can automate the process! Rather than maintaining and uploading classifications each time your tracking codes change, you can create automatic, rule-based classifications and apply them across multiple report suites using Classification Rule Builder. Rules are processed at frequent intervals, depending on your volume of classification related traffic. Rules can also be applied retroactively, so it's never too late to get them setup.

You can build rules with simple 'starts with', 'ends with', or 'contains' criteria, or get more advanced with regular expressions (RegEx). RegEx is helpful when the dimension is passing in a delimited value with a fixed format, e.g. Report Suite ID | Report Suite Name, where pipe is the delimiter. Learn more about how to build flexible rules using regEx.

Example RegEx ruleset, that will parse apart Report Suite ID | Report Suite Name into two classification dimensions

Rules can be tested before applying them to your full datasets as well, so you can have confidence in how they will process.

Sample test output

Classifications API

The Classifications API is another method for automating uploads. Learn more.

Upload Best Practices

Processing time for a classification file varies with the size of the file and the number of files that Adobe's servers are processing. Classification uploads typically take no longer than 72 hours to process. In order to prevent increased processing times, Adobe strongly recommends that you use the following guidelines.

  • Upload only changed rows (the “delta”)
  • Use the Classification Rule Builder when possible & choose “overwrite only unset values” option
  • Plan ahead – classifications can be uploaded before any hits for those values have been received. Keep in mind that uploaded values count toward a dimension’s uniques limit.
  • Upload small files <1MB via the browser importer & use FTP import for larger files
  • Combine classification files where possible into <200 MB sized files
  • Avoid uploading files over 500 MB
  • Avoid uploading large quantities of files to FTP, especially when there are many associated report suites
  • Additional best practice information
  • Troubleshooting uploads

Upcoming Improvements

There are several improvements coming to Classifications over the next few quarters. In September 2019, an update to the Classifications architecture began rolling out, with a small number of early adopters migrating to it. The update significantly reduces the time it takes for uploads (including rule logic) to be imported/ingested and made available for reporting. This improvement should not impact existing upload methods or workflows; the data is simply processed faster by Adobe.

More details about additional improvements, including a new Classifications API, will be shared as they become available.

In order to ensure your organization can adopt all of the improvements coming to Classifications, confirm the following:

  • Your organization has an established Experience Cloud Org
  • Your organization’s report suites have been mapped in the Experience Cloud. Learn more

Ready for more content?

Visit adobe.ly/aaresources for a full list of Adobe Analytics Spark pages & other helpful resources.

Created By
Jen Lasser