Adobe Experience Platform

Hi AEP Community, Greetings! How much time does it take for profile attributes to sync in the profile store when ingestion type is streaming?I had the idea that, all profile attributes that is either ingested as batch ingestion or streaming ingestion (from a true streaming source), then it takes about 24 hours or 24+ for the values to be synced in the profile store and only then we could create segments. Is my understanding correct? or does streaming ingestion helps with profile attributes being updates in under few minutes and the audiences would qualify in under few mins? Thanks,Sambit

Hi, I’m trying to setup a daily dataflow from AEP to an SFTP using Destinations in AEP. I’ve setup the SFTP destination and have been able to connect to it successfully, but keep getting the following error:Unable to create directory for path sftp/SFTP_220260324_120519/part-00000-tid-2776579317160816273-5bb6a03a-8b4a-46fc-8fca-d3f38be573b0-1414191-1-c000.json.gz with the parameters flowId=9a432588-4c13-442c-85c0-9b968c07a5c3 datasetId=65aefbff544b412c9f33e273 flowRunId=d9568853-f592-4361-9b08-df20bb1db29f. The error message returned by the destination is Can't make directory for path "sftp" under "/".. Please verify that you have the necessary permission to perform the operation to the destination and the path is correct. We’re aiming to have the file upload to the root folder under the SFTP, and have confirmed that AEP has all of the necessary permissions to read and write the root folder in SFTP, but are still encountering this error. Has anyone encountered a similar issue in the pas

Hi Team,We are observing several hundred batch IDs that have been stuck in a “loading” state in 2 months period, which suggests potential data loss.The source system successfully sent the data, however these batches never transitioned out of the loading state and appear to be stuck indefinitely. As a result, the data is not available downstream and we are currently missing critical data.We have subscribed to dataflow ingestion alerts, but these alerts are not providing any visibility into batches that remain in the loading state, and no updates are triggered for these batch IDs.Given that the data is now outdated and cannot be backfilled or re‑ingested, what is the most efficient and recommended approach to immediately identify, monitor, and detect batches that are stuck in the loading state?we are using Retrieve list of batches with query params so that we can see the missing batches that are still in loading state Any suggestions would be appreciated ThanksReshmika

Hi Team,We have three different data sources. One of them is Marketo. When Marketo deletes a record, we only see the isDeleted flag in the Data Lake, but the deletion is not reflected in the Profile Store.If the same profile later arrives from Source 2 or Source 3, it still qualifies for the segment and gets pushed back to Marketo through the Marketo Engage Destination Connector.Is there a way to prevent such deleted records from syncing back to Marketo again?Thanks in advance!

Adobe Tags does not currently allow for multiple development library configurations - are there any methods that individuals are using to allow for multiple individuals to work at the same time within the same property and whilst ensuring that the latest instance of the Adobe Tags? Is there Thanks,Ty

I am going to configure CNAME in order to have first party tracking. Is it also required to configure FPID? What is the specific use-case for FPID?https://experienceleague.adobe.com/en/docs/experience-platform/collection/use-cases/identity/first-party-device-ids

Hello Community - I'm hoping to get a better understanding of what makes "Tags" tick.   I have past experience with Adobe Tag Manager, DTM and now Tags. I have experience with GTM and Tealium. Tealium has a document called "order of operations" which explains internal prioritization of the load rules, extensions, tags, etc. I've not found the equivalent for "Tags" (or Launch).  The real question I have at this point is, are there some things which get cleaned out of Custom Code sections (either in events or rules) during the library build process?  I've added lines and lines of code with console.log statements and such and it simply disappears from the library or never gets executed (ignored?) when testing on a site.  Last, what's the story with trigger() ?  Is there any limit to where and when it can be used ?  Thanks for reading and I look forward to replies!

Hi,Created Data stream for AA [1] and another datastream for AT[2], now in the tag  properties updated AEP extension to set the datastream to [1].Within the rules, created one rule for AT - library loaded- send event - checked the include render propositions and checked the render visual personalization decisions with send display event - unchecked. The datastream override is done here to point to [2].Now for the AA rule - Only AA tracking is done -  on on load events - and unchecked the include render propositions and unchecked the render visual personalization decisions with send display event - unchecked. The datastream override is NOT done here and points to [1]. AA reporting seems to occur but Adobe target does not get the reporting . We have no a4t provisioned. @bjoern__koth 

Hello everyone,I am currently trying to understand how schema relationships work in Adobe Experience Platform. Specifically, I am looking at the example involving Loyalty Members (Profile class) and Hotels (Reference schema).My Scenario: I have established a relationship where the "Loyalty Members" schema (Source) points to the "Hotels" schema (Reference) via a preferredHotel field. Both schemas are enabled for Profile. I want to create an audience to segment all hotels where petsAllowed is True. My Question: When I export this audience to a Destination, can I export the attributes of the Loyalty Member profiles (likepersonalEmail.address or name) that are linked to those qualified hotels?If this is possible, how should I configure the Mapping in the Destination to ensure the Profile data is sent even though the segment criteria are based on the Reference Schema?References: Documentation Tutorial: Define a Relationship Between Two Schemas Using the Schema Editor | Adobe Experience

Could you please clarify whether merge policies (such as Dataset Precedence and Timestamp Ordered) are applicable to Account profiles in B2B implementations, similar to how they are applied for customer/person profiles in Adobe Experience Platform?Specifically, I would like to understand if merge policies control how data is prioritized or combined when multiple datasets contribute attributes to the same Account profile, just like they do for customer profiles.Any clarification or documentation reference would be greatly appreciated.

The Shopify Order_Line_Items table is missing a timestamp field needed to schedule a dataflow connection to pull in order details that relate order and products from Shopify into AEP. I was able to schedule data connections for the Orders and Customers tables. However, the Order_Line_Items table is needed to tie the products to the orders. Without this timestamp field we cannot use this Shopify connector. 

Hi,Has anyone experienced issues with mobile SDK events on iOS version >26.2?We are facing cases where events reach AEP for devices with version 26.2 or older, but not for newer versions in the same use case

I am getting this error when sending an audience to the TikTok Destination within AEP. Any suggestions?  {"message":"externalAudienceId could not be parsed","code":"500 INTERNAL_SERVER_ERROR"}

Hi AEP Community,I’m doing a POC, ingesting Experience Event data and Individual Profile data, into Adobe Experience Platform and I’m seeing this behavior: Data Lake ingestion works (5 records appear) Real-Time Customer Profile shows 0 records / 0 identity matches Here’s my setup: (Please note all are dummy data).Sample Dummy Profile Schema Payload:{ "_object": { "StCd": "CA", "CNUM": "1125" }, "_id": "/uri-reference", "personID": "Sample value"} Primary Identity: CNUM Namespace: CNUM Sample Dummy Event Schema Payload:{ "_object": { "EnEvenObj": { "CNUM": "1125", "Type": "Dep", "EventDate": "2026-02-20T20:20:39+00:00" } }, "eventType": "adve.comp", "timestamp": "2026-02-20T20:20:39+00:00", "identityMap": { "key": [ { "id": "Sample value", "authenticatedState": "ambiguous", "primary": false } ] }}Observations: Namespace and Primary Identity are defined in both schemas Data was ingested after Profile was enab

Hi all,We are using AEP B2B and Marketo as marketing automation which sends out the emails. We have the marketo person, marketo activity etc as AEP data source.We are receiving every click/ open etc events in the marketo activity source. Now we need to find the content which generated the marketo event. Can someone please shed some light on this, I am not able to find any column related to parameter and not sure if I am missing something here

Hi Team,We have successfully ingested data into the B2B Account and B2B Person datasets, and the profiles are visible in the Profile Store as expected. The profiles for these entities have been verified in the Profile UI and appear to be correct.However, after ingesting data into the B2B Account–Person Relationship dataset, we are unable to see the corresponding Account–Person relationship reflected in the Profile Store.For the relationship mapping, we followed the Adobe documentation provided in the link below:https://experienceleague.adobe.com/en/docs/experience-platform/sources/connectors/adobe-applications/mapping/salesforce#account-contact-relationBelow are the dataset details used in this implementation:Datasets: B2B Account_IM_Dataset B2BLead_IM_Dataset B2BContact_IM_Dataset B2B Account Person Relation_IM_Dataset Environment Details: Sandbox: aepenablementfy26 Environment Type: Development Region: VA7 Although the data ingestion for the Account–Person Relationship data

After sending incremental data to destination platforms (based on the referenced use case), we need to regularly delete the previous day’s batches from the destination dataset table. To automate this process, we explored the available API options, but it appears that the Batch Deletion API is deprecated.Could anyone confirm whether there is any supported method to programmatically delete batches?(Using the UI is not feasible since this needs to be fully automated.)Approaches We Explored but Could Not use:Deleting Records via Data Lifecycle Management - This does not work because the dataset is not Profile-enabled, so there is no primary identifier available to target records and we don’t want to make it enabled as well just to create duplicates in the profile store. Data Retention Policy - Retention policies apply only to event datasets, so they cannot be used for this dataset type.  

Hi all,I am using “Event grouping using session storage” feature to track activity map links in Adobe Analytics using Web SDK. (As shown below) For my Adobe Target and Adobe Analytics setup, I am using 2 send events on every page load.eventType:- decisioning.propositionFetch and "renderDecisions": true eventType :- Web Webpagedetails Page Views and "personalization": {     "defaultPersonalizationEnabled": false,     "includeRenderedPropositions": true   } so that the display event is sent for A4T purpose along with AdobeAnalytics variables (eVars, props and events).Due to this behavior, Activity Map details are getting attached on decisioning.propositionFetch server call on the subsequent page and not available in Adobe Analytics reporting. Is there any way to make the Activity Map to attach on to subsequent page view server call? I don’t want to use “No event grouping” as it increases server call usage.Thanks in advance!

Hi,Can we use lookup attributes in audience criteria? I am looking for support around using the Lookup class in Audience criteria. We’re validating our RTCDP modeling approach for campaign eligibility and would appreciate architectural confirmation. We were asked last minute to include divisionNumber in an audience eligibility which is apart of the lookup schema and dataset in dev sandbox.Current setup:Profile schema (XDM Individual Profile) identities: accountID, urlAccNameCard Lookup dataset contains divNoLogo (Division Number + Logo)Division Number = first 3 digits of divNoLogo (fixed length, 3 digits)Card Lookup does not contain accountID Card Lookup was previously profile enabled Requirement: segment based on Division NumberQuestion:Given that Card Lookup is not keyed by accountID, is Adobe’s recommended pattern to:1. Materialize divNoLogo (or derived divisionNumber) onto the Profile store via a profile enabled enrichment dataset keyed by accountID, and segment directly on the Pro

Dear community memebers,I’m sharing here a pretty odd experience we are facing with a specific data model passed through CSV files.The time-based dataset should import SFMC customer interactions with some columns mapped in SFMC data model as boolean but in numers (0 / 1).The oddity comes after the mapping of the original column to our XDM attribute, which is set as integer.After launching the import of simple test files, these attributes are not filled at all, as if there are no values in the source files. This is something we never faced, and a different column in the file mapped as number, is correctly populated.Any thoughts?

Hi! Greetings for the day. We are implementing CJA for a leading Healthcare client. We are currently in the discovery phase, where we propose that they start with the CNAME approach using the Web SDK. Given that the CNAME approach is also not full-proof with browsers like Safari, they also want us to explore the Edge Network API for a future-ready implementation. They don't have CDP, and for now, they will only be sending web events (website interactions)-related data to AEP. What I understood is that the API is useful for environments that do not support Web or Mobile SDK, but we can deploy the Web SDK JS on their website. They are considering the advantage of a secure and authenticated way to send sensitive data to the platform via the API, but they only have the website/clickstream data at the moment, which I believe will go through the unauthenticated endpoint (or can we use the authenticated endpoint here?). I believe sending events data for 1000s of website users via the API is n

Hi, I am having a field which is present in the Union schema and it has values for a number of records. I can see it all right in the profile:Profile/Browsein Union Schema:Union SchemaBut if I run this query on the profile snapshot:SELECT to_json(_myTenantName)FROM profile_snapshot_export_66aefd28_7113_489a_a40f_b7fa9efe34e5 AS pWHERE p._deloittellpcanada.partyIdentifier.CID= '111613'I get this - no SuppressionID_List:Any idea why is that and how can I get all fields into the Profile Snapshot?

Hello,I am currently setting up a source flow from Salesforce CRM. I am following this guide to filter ingestion. https://experienceleague.adobe.com/en/docs/experience-platform/sources/api-tutorials/filter#specify-filtering-conditions-for-ingestion I get an error message when the flow runs. The API request to Salesforce failed. Request Url: /services/data/v59.0/query?q=SELECT+FirstName+FROM+Contact+WHERE+`FirstName`+LIKE+'Ell%25'+AND+(LastModifiedDate+<+2026-03-04T04%3a45%3a00Z), Status Code: BadRequest, Error message: [{"message":"\nSELECT FirstName FROM Contact WHERE `FirstName` LIKE 'Ell%' AND (LastModifiedDate\n ^\nERROR at Row:1:Column:36\nunexpected token: '`'","errorCode":"MALFORMED_QUERY"}] This error message could be because of the ` character used in the query. `FirstName`Salesforce does not accept ` characters.The connection spec for Salesforce accepts filters but seems to have the "columnNameEscapeChar" set to ` which is probably not correct?  ```"filterAtSource": {     

Hi everyone,I created a Customer AI instance where the goal is set to a specific event type. When I try to run the instance, it fails with the following error:Error Code: VALIDATION-400 NOT_ENOUGH_OBJECTIVERequire at least 50 users with qualifying events to build a model.I already have more than 50 users meeting the qualifying event criteria, but the instance is still not generating scores.Could someone please clarify what the actual minimum number of users required is to successfully generate scores? Also, are there any additional constraints or thresholds (such as training window requirements or minimum events per user) that must be met?Thanks in advance!

Hey, I’d have a need to exclude some attributes going from the hub to edge (profile). Suggested approach is to use data governance labels (https://experienceleague.adobe.com/en/docs/experience-platform/data-governance/labels/reference). None of the ootb labels seem to be related to the edge case specifically. Anyone who have done this exercise and would know what labels to use?Attributes should still be usable for personalization cases that rely on data in the hub.