Request: Migration guide for moving from FTP classification importer to Classification API

[upliftertom]

We're an ISV that provides a service for our enterprise clients that use Adobe Analytics to update their saint classifications amongst other services.

As the FTP importer is due to be deprecated on 31 August 2026, I know for us at least we'd really benefit from a migration article if at all possible.

In particular we would like to know:

• The scope to request to add to a server-to-server api key to access classification api
• What we would need to tell the client to prepare for the change to classification sets, and how it alters the file to upload

Additionally, the documentation for file import on this page:
https://developer.adobe.com/analytics-apis/docs/2.0/guides/endpoints/classifications/import-file#post-import-upload-file

References this page for the file import:
https://experienceleague.adobe.com/en/docs/analytics/components/classifications/legacy-classifications/classifications-importer/c-saint-data-files

Is that actually correct? With the processing header still?

[johnwight]

@upliftertom thanks for your question. Let me make sure I understand your request. You would like more detailed instructions for:

1. Migrating from SAINT to AA 2.0 Classifications API
2. Recommended scopes in 2.0 API requests for Classifications
3. List of articles explaining classification sets preparation (I think we have this already on Experience League, but good to aggregate and synthesize the info).
4. I'm not sure I understand the question on the "processing header" for importing files with the 2.0 classifications API. I would not use the legacy instructions when forming your requests with the 2.0 API. Or are you asking about how to continue importing with SAINT as you are currently doing? Please let me know your context.

Does this encapsulate your inquiry? Thanks.

[upliftertom]

Yes to 1, 2 and 3 - SAINT has been around a long time and these changes are not insignificant.

For 4- the processing header is this bit (from https://experienceleague.adobe.com/en/docs/analytics/components/classifications/legacy-classifications/classifications-importer/c-saint-data-file)

The instructions in the article https://developer.adobe.com/analytics-apis/docs/2.0/guides/endpoints/classifications/import-file#post-import-upload-file seems to refer to classification sets, but that link references the legacy classification format for the upload (the link is where I highlighted "classification data files" in my first comment)

[johnwight]

@upliftertom thank you for responding. I think that link looks technically out of date for the API and is therefore just confusing. I will remove that link to the Experience League article. I will also aggregate and assess the resources available for further guidance on your migration and confer with the team on how best to address your need. I'll report back. Thanks.

[upliftertom]

Thanks John

One further question related to migrating, we went through with the client and saw that there were classification sets set up already that matched what we were uploading for our SAINT campaign classifications.

We will probably ask the client to reach out to adobe support, but we noticed that for one of the classification sets the "classification name" wasn't related to the "identity name". (the one set for "Source" was set to an identity name of "publisher1" for example)

What is an identity name? is it the name of dimension for the api? the docs on:

https://experienceleague.adobe.com/en/docs/analytics/components/classifications/sets/set/schema

Say:

The part "the the" makes me think something got cut off

[upliftertom]

Also I think I found some further typos:

https://developer.adobe.com/analytics-apis/docs/2.0/guides/endpoints/classifications/import-file#post-import-upload-file
and
https://developer.adobe.com/analytics-apis/docs/2.0/guides/endpoints/classifications/import-file#post-import-commit-job-id

they have

https://analytics.adobe.io.api/

which I suspect should be

https://analytics.adobe.io/api/

and the example has "UTF8" as the value for encoding but the table has "The encoding for data. The default value is UTF-8."

[johnwight]

Hi @upliftertom, Thanks very much for pointing out those URI path and utf8 typos. The developer who can clarify with an actual current response is on vacation now, but I will leave a note for him.

Also, your query has shown that I need to update some links in the API documentation to Experience League. I discussed your issue with the team and it appears that the previous link did not send you to the right Experience League article that is meant to help with migration for SAINT: https://experienceleague.adobe.com/en/docs/analytics/components/classifications/sets/overview

They asked that you follow that guide and provide feedback of any gaps in info to them through this repo issue. Although this repo is meant for developer.adobe.com documentation (strictly APIs), you can use this to have me get it to them (those who document Experience League and UI conceptual topics).

Regarding "identity name" in the Experience League documentation, I will ask @ridmaur to help out with that understanding.

[ridmaur]

Updated the Experience League doc on Identity name to correctly explain why it is there and how it is used.
https://experienceleague.adobe.com/en/docs/analytics/components/classifications/sets/set/schema

[upliftertom]

Updated the Experience League doc on Identity name to correctly explain why it is there and how it is used. https://experienceleague.adobe.com/en/docs/analytics/components/classifications/sets/set/schema

@ridmaur

We noticed for one of our clients that the classification name and the identity names seemed unrelated on one of the classification sets that had been created automatically (which we learned happened through this video https://www.feld-m.de/en/events/on-demand-recording-road-to-summit-2026/ and isn't mentioned in the docs anywhere) - a "Source" classification name had an identity name of "publisher" for example. I suspect it was the "initial value" but is there any practical impact that could confuse users and affect reporting?

I'm still not clear on the purpose of the identity name- it seems like this is the only place it is shown and it doesn't seem to have any impact or use in the API - is it something Adobe uses internally (would I ever use it?) In reporting wouldn't clients be breaking down by the classification name?

(edit)

ie would we be better deleting and recreating this classification set?

--

I think I figured it out- it's the name of the variable in the API:

https://analytics.adobe.io/api/:globID/dimensions?...

Is it used anywhere else?

[ridmaur]

@upliftertom As mentioned in the updated docs: "Identitfy name is the system generated name for the classification. This name is a read-only value and is derived from the initial classification name. The identity name is used internally to persist the classification, act as the ID of the reportable dimension, and cannot be changed." There is no other use for the identity name other than for internal purpose.

[upliftertom]

Great, thanks for your help - I have no further questions.

We've completed our migration, and I wanted to say that this new method is much better. I particularly like the notification framework if the upload fails. And the format of file uploads- glad to be away from the eccentricities of the original SAINT import format!