rowsXcols

CJA Readiness: Three key considerations before migrating to CJA

Migrating from Adobe Analytics (AA) to Customer Journey Analytics (CJA) is a complex but valuable transformation that enables businesses to leverage more advanced analytics capabilities within the Adobe Experience Platform (AEP). The pre-migration process depends on several factors, primarily focusing on data collection, the existing Adobe Analytics setup, and custom integrations.

In this guide, we will explore the three mentioned key considerations to ensure a smooth migration planning process—or as we call it, the CJA readiness stage.

1. Understanding data collection requirements

The importance of data quality

“Garbage in, garbage out.” Ensuring high-quality data collection is essential, as it forms the foundation of your analytics. A thorough review of your tracking implementation is necessary before migration to ensure accuracy and consistency.

Web SDK vs. AppMeasurement

One of the most critical aspects of migration is assessing the current data collection setup:

  • If your platforms (Adobe Tags properties) already run on Web SDK, the migration will be more straightforward.
  • If your platforms still use AppMeasurement, additional time will be required to adapt, as Web SDK introduces several new concepts, such as Experience Data Model (XDM) schemas, identities, and datasets. While AppMeasurement can technically be used with AEP, it introduces additional complexity over time. I recommend strongly to use WEB SDK only.

Reviewing the data layer and tag management system

Migration presents an opportunity to revisit and optimize your data collection approach:

  • Align and standardize the data layer across different platforms. Choose your right data layer setup. May update your data layer from good old “Customer Experience Digital Data Layer” (CEDDL) to an event driven (EDDL) or a hybrid-one.
  • Ensure Adobe Tags (Launch) or any other tag management system is optimized to support AEP/CJA requirements.
  • Review the Solution Design Reference (SDR) and align data collection strategies to meet CJA requirements.

Our Approach

Fortunately, we had already migrated all our platforms to Web SDK and were familiar with AEP concepts. Additionally, our data layer and tag management setup were standardized across all platforms (we are using a hybrid data layer approach, a combination between CEDDL and EDDL). Nonetheless, we conducted a thorough audit of our launch properties and SDR. We ensured that key attributes, such as page and event data, were tracked consistently with high data quality. Within the SDR, we critically assessed every attribute—questioning its necessity and evaluating how it could be improved using CJA’s new capabilities (compontent configuration possibilities, e.g. derived fields).

2. Evaluating your Adobe Analytics setup

Your current Adobe Analytics environment plays a significant role in the complexity of migration. Key considerations include:

Data migration strategy

When migrating data from Adobe Analytics to CJA, it is essential to carefully determine what data should be migrated and the appropriate time period (backfill length). Instead of transferring everything, use this opportunity to refine your analytics setup and tracking plans, ensuring only relevant data is included.

By default, Adobe allows for 13 months of historical data import into CJA. HBy default, Adobe allows for 13 months of historical data import into CJA. However, depending on your business needs, a longer data retention period may be necessary. For example:

  • If your business experiences seasonal peaks (e.g., September to November) and operates on a three-month analysis cycle, you may need 15 months of historical data. This consideration is important not only for analytical purposes but also for licensing requirements.
  • A longer data retention period allows for better year-over-year comparisons and trend analysis, but it also increases data volume, storage costs, and processing complexity. Carefully evaluate the use cases you want to cover with CJA.

Balancing data retention needs with storage considerations is crucial for optimizing your CJA setup.

Choosing a data migration method

Deciding how to transfer your data to CJA is another crucial step. There are two primary options:

  • Adobe Analytics Source Connector: A simpler, more automated method for integrating with CJA.
  • Data Feeds: A more flexible but complex approach, allowing deeper customization of data transfer.

Choosing the right method depends on your specific data needs and infrastructure. For more some lessons learned regarding data migration, refer to this article.

Component migration

Rather than migrating components one-to-one from AA to CJA, this transition presents an opportunity to start fresh. Over time, Adobe Analytics implementations often accumulate redundant, outdated, or poorly documented components.

Our Approach

We avoided using the Component Migration Tool and instead built a new, streamlined setup. To ensure a smooth transition, we conducted a stakeholder analysis to determine which dashboards were truly necessary, reducing their total number by at least 50% and eliminating duplicated or unused reports and components. We reviewed and refined segments, metrics, and other components to prevent legacy elements from being carried over.

For data migration, we opted for data feeds rather than the Adobe Source Connector due to its limitations (we don’t wan’t any eVars and props in our fresh CJA setup). Rather than simply transferring old complexities into the new system, we treated the migration as an opportunity for cleanup and optimization, ultimately creating a more efficient analytics environment (also boosting self service analytics).

3. Custom integrations and data transformation

This is often the most challenging part of migration. Many organizations integrate Adobe Analytics with third-party integrations such as:

  • Data warehouses (via API, FTP, or custom pipelines)
  • Mailing systems, marketing automation tools and recommendation systems
  • CRM and personalization engines

Since CJA operates within AEP (and has some limitation regarding export), these integrations must be reconfigured using available options, including:

  • AEP Data Ingestion APIs
  • Adobe-built and custom-built connectors
  • Data Prep Pipelines for transforming and routing data

Data transformation challenges

Data transformation is a major challenge during migration. While standard connectors provide some level of transformation, API-based approaches (e.g., Query Service) require careful handling of object-oriented AEP data to relational structures (e.g., tables, views, or data lakes). Properly structuring and optimizing these processes is essential to ensure data usability across different platforms.

Our Approach

Our data import and export setup was relatively straightforward, though we did transfer some data into our internal data lake. For this, we relied on daily Data Warehouse exports via FTP and the Data Warehouse API. Since CJA currently has limited options for such exports (e.g. full table export support 10 dimensions and 10 metrics), we chose to export data on a dataset basis from AEP.

For our needs, the Query Service API combined with AEPP proved to be the most effective approach. This allowed us to access datasets from our internal data lake and persist them as needed. However, since the data originated from AEP rather than CJA, it lacked persisted attributes such as last-click attribution or visit-based metrics. To bridge this gap, we used SQL and Python to recreate these elements. Fortunately, Adobe provides predefined functions for visit identification, and standard SQL window functions make it possible to reconstruct everything available in CJA.

Planning data pipelines in advance is crucial, as modifying these processes requires internal IT resources. The more import/export operations involved, the greater the complexity, increasing both maintenance effort and resource demands. Keeping the process as streamlined as possible helps minimize overhead while ensuring data consistency.

Final thoughts

Migrating from Adobe Analytics to Customer Journey Analytics is not a simple lift-and-shift process—it requires thoughtful planning, data optimization, and strategic decision-making. By reviewing data collection, refining components, and carefully managing integrations, businesses can unlock the full potential of CJA while avoiding unnecessary complexity.

A successful migration lays the foundation for a more powerful, flexible, and future-proof analytics environment within AEP.

In