AEP implementation: an end-to-end guide

Step-by-step implementation

1. Prioritise use cases — Agree 2–3 measurable use cases with business owners and define success metrics.
2. Design data and identity — Create the XDM model and identity strategy needed for those use cases only.
3. Set up ingestion — Configure sources, Data Prep mappings and Web SDK collection in a development sandbox.
4. Build profiles and audiences — Enable datasets for Profile, choose merge policies and build validated audiences.
5. Activate — Connect destinations or Journey Optimizer and validate counts and consent.
6. Test and go live — Run end-to-end QA, promote to production, and set up monitoring and ownership.

Implementation checklist

• Use cases, KPIs and owners documented.
• Source inventory with data quality assessment.
• Schemas reviewed; identity namespaces created.
• Consent and governance labels applied.
• Sample data validated in profiles before full loads.
• Audience definitions signed off by marketing.
• Monitoring, alerting and runbooks in place.

Project roles

Typical roles are a solution architect, data engineer, implementation specialist (Web SDK and Tags), marketing technologist for audiences and journeys, and a business owner. Smaller teams combine roles; enterprise teams add data governance and QA leads.

Common risks

• Starting with every data source instead of one use case.
• Identity decisions made late, causing rework.
• No owner for data quality after go-live.

Who this guide is for

This guide is written for beginners, students and career switchers who want a clear mental model, and for developers, data engineers, implementation specialists, consultants, architects, marketers, agencies and enterprise teams who need practical, step-by-step detail they can apply on real projects.