[Mentor Jayakrishnaa Parthasarathy] Adobe Real-Time CDP Community Mentorship Program 2024

This mentorship program sounds fantastic! I'm excited to learn from Jayakrishnaa Parthasarathy and connect with fellow aspirants. His expertise with Adobe platforms is impressive. Looking forward to a great learning experience and the chance to win some Adobe swag.

Hi Team,

Hope everyone had a good weekend. Please find the below quiz/questions that is related to XDM concepts. Feel free to post your understanding on the quiz and let me know if any. I have created few questions based on the document shared earlier on this thread. Will be adding few more shortly.

1. What are the ideal steps to convert any normalized sources of data to AEP XDM?
2. Is it really required to consider the secondary identifier for the entity? What is the actual purpose of having secondary identifier on the schema?
3. Under what circumstance we should be considering the custom class? i.e., look up schema.
4. Is it necessary to have email attribute as “Email” data type or just a string should be fine? What is the difference on using email & string as a data type for email attribute?
5. Sources like CRM, Loyalty, Member entity.etc. with respective identities. What is the identity namespace type that would be considered for the respective identities for the mentioned entities?
6. Give me some examples of source/entity name where we will use “non-people identifier” as primary identifier? What is the main objective of considering “non-people identifier” as primary identifier?

Thank you,

Jayakrishnaa P.

Hi Team,

Good afteroon. Please find the quiz for the week and try to give your answer. Feel free to comment if any. 

1. When two profiles are merged due to having the same email ID in the identity service, how can you identify the datasets and batches responsible for this merge?
2. If profile fragments ingested with an email as an identity are not appearing in the identity graph and the system returns "No identity graphs found," what could be the potential reasons for this issue?
3. Which feature in Adobe Experience Platform allows you to view the union of specific class fields, attributes, and the corresponding schemas and datasets?
4. Is it possible to delete profiles that have been ingested into Adobe Experience Platform? If so, what methods can be utilized to remove these profiles from all three services?
5. What troubleshooting steps should be taken if batch data has been successfully ingested into Adobe Experience Platform, but the profiles are not visible in the profile tab?
6. What are the potential impacts on the system if data is ingested with a data type mismatch, such as specifying "String" during data modeling but receiving "Integer" data?
7. What is the maximum number of identities that Adobe Experience Platform can hold for an individual profile, and what happens if this limit is exceeded?
8. What are the consequences of deleting a dataset that has been enabled for profiles in Adobe Experience Platform?
9. What are some common challenges encountered during data ingestion in Adobe Experience Platform, and how can these issues be mitigated?

Happy Learning!

Jayakrishnaa P.

Hey @jayakrishnaaparthasarathy , Everyone,

Here are few questions, please feel free to answer

1. What is the difference between Field group and Data types?
2. How does AEP handle privacy and compliance requirements within the context of profile and identity management?
3. What are the best practices for handling large volumes of data ingestion into Adobe Experience Platform to ensure optimal performance and data quality?

1. Audience Configuration: When creating an audience, select the appropriate settings like "show only fields with data," "show full XDM schema," and "merge policy." Enable the active edge service to evaluate the audience on the edge network.
2. Query Type Selection: Choose the correct query type, such as those enabled for streaming segmentation, to ensure the audience is grouped based on the relevant events.
3. On-Demand Evaluation: Use the on-demand API payload to check profile counts whenever needed.
4. Audience Maintenance: Audiences with membership in the exited status for over 30 days, based on the last qualification time, may be removed from the audience list.
5. Real-Time Segmentation: Streaming segmentation qualifies profiles as soon as data arrives, eliminating the need for manual scheduling. Ensure a well-structured data flow for accurate and timely segment evaluation.
6. Complex Segment Rules: Utilize combinations of conditions (AND/OR logic) and nested groups to create advanced audience segments targeting specific behaviors and characteristics.

1. Review Before Certification: Given the variety of audience attributes and evaluation methods, review the entire category before attempting certification. Reference: Link.
2. Attribute Mapping: Ensure supported attributes are selected during the mapping step to export profiles correctly. Missing attributes may prevent profile export to destinations.
3. Scheduled Activation: Schedule activation based on the use case to ensure profiles are delivered to the destination on time for marketing campaigns.
4. Follow Activation Guardrails: Adhere to activation guardrails to avoid errors. Reference: Guardrails.
5. Data Mapping Consistency: Maintain consistent data mapping between your platform and the destination to ensure accurate data transfer and avoid errors or incomplete activation.

Happy Learning,
Jayakrishnaa P.

1. What are the different methods of audience creation in Adobe Experience Platform, and what is the core difference between these methods?

Ans:     

A. Segment Builder:

The Segment Builder is primarily used for audience creation, allowing users to build audiences from scratch. 

B. Audience Composition:

It is used to create new audiences based on existing audiences,

Addition to audience compositions and segment builder, we  can use Adobe Federated Audience Composition to build new audiences from enterprise datasets without copying underlying data and store those audiences in Adobe Experience Platform Audience Portal.

2. If you’ve ingested batch data from S3 and want to combine it with streaming attributes for audience creation, can this audience be evaluated using streaming evaluation? If yes, how does the evaluation process work?

Ans:

AEP supports the evaluation of audiences using streaming segmentation, even if the data has been ingested from source like S3.

Streaming segmentation is an ongoing data selection process that updates your audiences in response to user activity. Once an audience has been built and saved, the segment definition is applied against incoming data to Real-Time Customer Profile. Additions and removals to the audience are processed regularly, ensuring our target audience remains relevant.

3. Can you provide an example scenario where using the audience composition method is preferable over a build rule

Ans:

Business Context:

A retail brand runs a loyalty program and wants to create a segmented audience for a special promotion. The criteria for the promotion are as follows:

1. High-Value Customers: Customers who have spent more than $1,000 in the last year.
2. Engaged Customers: Customers who have made at least three purchases in the last six months.
3. Recent Visitors: Customers who have visited the website or opened the app within the last 30 days.

The brand wants to create an audience that includes customers who meet any of these criteria to maximize reach for the promotion.

Using Audience Composition:

• Step 1: Create individual audiences for each criterion:
• Audience A: High-Value Customers
• Audience B: Engaged Customers
• Audience C: Recent Visitors
• Step 2: Use audience composition to combine these audiences. In AEP, audience composition allows you to create a new audience by combining existing audiences using simple logic (e.g., Audience A OR Audience B OR Audience C).

4. After creating a schema and dataset enabled for a profile, why might the schema’s attributes not appear during audience creation if the data hasn’t been ingested yet? Which setting must be enabled to resolve this issue?

Ans:

5. How long does it typically take for a segment definition to become available on the Edge Network after it has been created?

Ans :

After a segment definition has been created, it typically takes up to one hour for the segment to become available on the Edge Network. 

6. If you’ve set up a batch evaluation audience with a 24-hour schedule, is it possible to estimate the profile count before the evaluation runs? If yes, how would you do it?

Ans:

it is possible to estimate the profile count before the evaluation runs for a batch evaluation audience with a 24-hour schedule using View Last Sample Status API get request.

7. What happens if you don’t add a mandatory attribute in the mapping section when activating a Marketo destination?

Ans:

When we don’t add a mandatory attribute in the mapping section when activating a Marketo destination, it can lead to issues with the successful export of data.

8. What identities are supported for activating a Bing destination?

Ans:  MAID – Microsoft Advertising ID

9. What are the prerequisites for enabling audiences in personalization destinations like Adobe Target or custom personalization?

Ans:

A. We must have at least one audience created in Adobe Experience Platform. This ensures that you have the necessary access and permissions to activate audiences and enable the mapping step of the workflow.

B. To activate audiences and enable the mapping step of the workflow, we need the following access control permissions.

Destinations, Activate Destinations, View Profiles, and View Segments.

C. When configuring the Adobe Target connection without using a datastream ID, it's important to note that only next-session personalization use cases are supported in the absence of a datastream.

10. During the Identity Mapping step, how does Adobe Experience Platform handle unhashed attributes, and what is the purpose of the "Apply transformation" option in this context?

Ans:

AEP handles unhashed attributes by providing the "Apply transformation" option, which automatically hashes the source attributes on activation. This option serves the purpose of transforming unhashed attributes into the required hashed format for seamless integration with the destination platforms.

• A) To classify data for organizational purposes
• B) To manage and enforce data usage policies
• C) To improve data processing speed
• D) To categorize data for better visualization

• A) The label is only applied to the dataset currently in use.
• B) The label propagates to all datasets using that schema.
• C) The label applies only to the fields currently visible in the schema editor.
• D) The label is applied manually to each dataset and field individually.

• A) Use of a 'POST' request without any additional parameters.
• B) Use of the 'If-Match' header to prevent data collisions.
• C) Application of a new schema before updating labels.
• D) Deletion of all existing labels before adding new ones.

• A) That the applied labels conflict with an existing data usage policy.
• B) That the dataset contains errors that need correction.
• C) That the schema is not correctly linked to the datasets.
• D) That a required label is missing from the schema.

• A) Role-based access control
• B) Attribute-based access control
• C) Password-based access control
• D) Token-based access control

• A) Data Export Tool
• B) Configuration Sync Tool
• C) Sandbox Tooling
• D) Role Management Tool

• A) Sandboxes are unrelated to permissions
• B) Permissions can be configured separately for each sandbox
• C) Sandboxes automatically inherit global permissions
• D) Only admin roles can access sandboxes

• A) Assign them a global admin role
• B) Add their account to the sandbox through the Permissions UI
• C) Modify their user account settings manually
• D) Create a custom permission for the sandbox