Adobe Analytics classifications are a method for grouping tracking data, enriching it with descriptive attributes, and performing group-level analyses.
What is the purpose and benefit?
Once you’ve created classifications, you can run analyses based on them. For example, instead of analyzing just a single product, you can also generate cross-product reports by product category, product color, product size, etc.
The previous Classification Rules were a relic of the past. For some time now, Adobe has been promoting its follow-up feature, “Classification Sets,” which is designed to combine the previous “Rule Builder” and “Classification Importer” tools into a single interface. However, we discovered a few bugs in the interface during our testing.
The old Classification Rule Builder will be discontinued at the end of February 2027, and the Classification Importer will be discontinued as early as the end of August 2026. By then, we should all have thoroughly familiarized ourselves with Classification Sets and, ideally, switched over to them. Classification Sets are the new standard in Adobe Analytics, optimized for the Adobe Experience Platform (AEP) and Customer Journey Analytics (CJA).
Previously, setting up classifications involved several steps. First, you had to go to the Report Suite Settings in the Admin area and define Classification Items for a dimension there.
Afterward, you could either use the Classification Importer to upload classifications for each Report Suite based on so-called SAINT files, or alternatively, set up rules in the Classification Rule Builder (in most cases using regex) according to which the dimension was classified.
Using regular expression (Regex) classification, it is possible to classify dynamic values that follow the same structure without having to upload the exact matching classification value for every instance of a value. A common example of this is tracking codes for campaign tracking.
All of these steps are now consolidated in one place within the Classification Sets. This is definitely a huge help as it makes creating classifications easier and clearer.
Previously, you could select “Report Suites,” “Classification Importer,” and “Classification Rule Builder” in the “Admin” tab. This will no longer be available, and the “Classification Sets” option will be available in the “Components” tab.
There you’ll find three sections:
Let’s dive right into the “Classification Sets” section. There, you’ll find sets that have already been created—drawn from the existing Rule Builder sets—which you can edit and delete, or you can create a new set by clicking “New.”
A set has the following tabs:
In the Settings, you can enter a name, a description (optional), and tags (optional) for the classification set. You can also specify who should be notified if the classification process is successful or unsuccessful.
Below that comes the important part: the so-called subscriptions. Here, you specify which variables from which report suite should be classified.
Subscriptions offer significantly more flexibility than the old setup, since you can select not only different report suites but also different variables.
Applying the same classification to different evars or props was not possible before. Also note: If you select a variable that is already classified in another classification set, you’ll receive an error message.
Let’s take a look at the other two configuration tabs:
In the "Schema" tab, you can upload files for classification, similar to the previous SAINT files.
First, download a template, fill it out, and then upload it along with the classified data.
The file looks a little different than before and now appears much cleaner. Instead of the header that previously had to be included in every SAINT file, only column names are now required – first “Key” (for the values to be classified) and then the names of the respective classifications.
In the "Rules" tab, you can set up classification rules, similar to how you did in the Classification Rule Builder.
Rules are useful when dimension items change constantly (in campaigns, for example) but there are clear rules governing how they are structured.
As in the Legacy Rule Builder, rules can be set up here as regular expressions (Regex) – probably the most common use case – or using other functions such as “Starts with,” “Ends with,” “Contains,” “Matches,” etc.
Unlike the old Rule Builder, all rules are set directly in the Classification Set and then apply to all defined subscriptions.
Getting comfortable with Classification Sets in Adobe Analytics
If you want to dive deeper: Fellow FELD Mie Eliot Jones explains how classification sets work and how they're most effectively used for Adobe Analytics setups in his on-demand session “Demystifying Classification Sets in Adobe Analytics”, recorded at FELD M's The Road to Summit 2026.
In the recording you’ll learn:
- how classification sets work,
- how they differ from the old classification system,
- and how to set up your classifications so they remain maintainable and analyzable in the long term.
Watch on-demand now: Demystifying Classification Sets in Adobe Analytics
Classification Sets are a useful and positive advancement that significantly simplify classifications and offer us more options for configuration and data analysis.
The entire configuration is located in one place, eliminating the need for cumbersome configuration across different parts of the admin area. We also appreciate the additional options, such as the ability to configure various variables within a single classification set.
Unfortunately, we encountered a few bugs during our testing, which we hope will be fixed soon:
For several customers, rules stopped working after migrating from the old Classification Rule Builder to Classification Sets. This was likely due to a difference in regex syntax. It’s important to keep this in mind, especially if you plan to migrate a large number of rules.
Another issue with the migration was that – based on our experience – all rules previously created in the system were automatically transferred to Classification Sets.
A separate Classification Set was created for each Classification Rule Set (i.e., per eVar) and per Report Suite. This means that the individual classifications are still contained within a single set, but during the automatic migration to the new interface, they were split up per Report Suite (a separation that does not exist in the legacy Rule Builder).
However, when a new Classification Set is created in the Classification Set interface, you can immediately select all Report Suites to which the classification should apply, without needing multiple sets for this purpose.
The automatic migration by Adobe resulted in a huge number of new classification sets for some of our customers. In addition, the owners of the rules received email notifications for each set, which in some cases led to hundreds of emails per day.
Therefore, it makes sense to first clarify any questions or uncertainties with the Adobe account administrator before the migration.
In general, we recommend using the transition to the new Rule Builder within the classification sets (scheduled by Adobe for February 2027) as an opportunity to review all active rules created to date and clean them up as needed.
You can then start fresh in the Classification Set, either by migrating the existing rules from the legacy Rule Builder or by recreating the desired rules in the new interface.