Adobe Experience Platform (AEP) is a powerful tool for managing and delivering personalized customer experiences. However, implementing AEP can be complex, and without proper planning, you may encounter common pitfalls that will seriously hinder your success. In this blog post, we'll explore the essential considerations for AEP implementations, specifically organization, standardization, and documentation.
Organize, Standardize, and Document
To maximize the potential of Adobe Experience Platform (AEP), it's crucial to establish a solid foundation in 3 areas:
Organization: Establish a clear ownership structure and logical grouping of assets to ensure easy access and management.
Standardization: Develop consistent naming conventions, data formats, and metadata standards to avoid confusion and errors.
Documentation: Maintain detailed records of integrations, datasets, segments, and journeys to facilitate knowledge sharing and troubleshooting.
Use Case Documentation
AEP use cases can be summarized based on key metrics:
Timeliness: Ensure real-time data processing and analytics to respond promptly to customer interactions.
Throughput: Optimize data ingestion and processing to handle large volumes of data without performance degradation.
Lineage: Track data origin, transformation, and movement to maintain data quality and integrity.
KPIs: Define and monitor key performance indicators to measure business outcomes and optimize experiences. In other words, the reason why you're implementing the Use Case.
Most of the Use Cases implemented in AEP follow a standard flow, that could be broken down in 5 steps:
Data Ingestion (e.g., Azure Blob Connector)
Data Processing (e.g., Adobe Experience Platform)
Data Storage (e.g., Adobe Experience Platform, Azure Blob Storage)
Data Activation (e.g., Segments, Destinations, Journeys)
Customer Experience Delivery (e.g., Personalized Content, Offers)
The important thing is to agree on a documentation format and follow it across. Consistency is key for people to know what to expect and search when looking at a document.
Data Naming
One of the things that can get very quickly out of control is the naming of components inside the platform. AEP does not enforce anything, so it is up to the implementer to decide the format. The best practice in this case is to establish consistent naming conventions beforehand for:
Schemas: Use descriptive names with versioning (e.g., "customer_data_v2")
Datasets: Include data source and type (e.g., "customer_purchase_data_azure_blob")
Segments: Use clear descriptions and target audience (e.g., "high_value_customers_eu")
Destinations: Identify platform and purpose (e.g., "facebook_ads_targeting")
Journeys: Describe customer experience flow (e.g., "welcome_series_email_journey")
Never, ever forget to add descriptions to everything you add to AEP. I lost count of the times that my past self has helped me avoid "what is this? Where this comes from?" moments.
Data flows
It is critical to document integration points and dependent datasets/activations, covering at very least:
Integration points, such as Azure Blob Connector, and dependent datasets and activations.
Clear definition of ownership and access controls, like customer-managed passcodes and IT-managed credentials.
Implement rotation schedules for keys and credentials, otherwise data will stop flowing into AEP and debugging can take a bit of time to realize that is just an expired token.
Identify and document developer Adobe IO accounts and non-expiring query credentials. I can't recall how many times Adobe IO Projects were deleted or tokens refreshed by unaware developers.
This is my first set of common pitfalls and best practices when using AEP. If you follow these, you'll be well on your way to unlocking the full potential of Adobe Experience Platform and delivering exceptional customer experiences.
Comments