Innovative Marketing Practices on the Cloud: How an Intelligent Data Lake Enables Flexible and Efficient Marketing Capabilities

This article is an excerpt from Troy Cui’s talk at the 8th GIAC Global Internet Architecture Conference, where he shared the case study "Innovative Marketing Practices on the Cloud – How an Intelligent Data Lake Supports Flexible and Efficient Marketing Capabilities".

Troy Cui, a data and technology platform architect with extensive experience in data products, middleware, and high‑performance deep‑learning platforms, now leads data analytics product promotion for Amazon Web Services China.

The talk outlines the background of the data‑lake‑warehouse (湖仓) concept, emphasizing the need for a solid data foundation to empower marketing scenarios. It presents consumer behavior trends from 2017‑2021, highlighting increased willingness to pay for value, convenience, and experience.

It then discusses the growing volume of data versus the complex demands of marketing, noting that enterprises must transform data into a core asset and address challenges such as data silos, insufficient data processing capabilities, and low analyst participation.

The core solution is the "Intelligent Data Lake" architecture, which unifies data governance, provides secure S3‑based storage, and offers specialized analytics services for high performance. The architecture avoids monolithic products, instead delivering flexible, extensible, and reliable data pipelines.

Key capabilities include:

Unified governance foundation covering data quality, permissions, and visualization.

Production‑grade data processing for machine learning, featuring open heterogeneous engines, elastic platforms, and data‑quality optimization.

Intelligent analytics that bridge technology and business value, enabling rich scenario support, model feedback, and business‑driven innovation.

Practical marketing use cases illustrate how multiple data sources (relational, NoSQL, object storage) are federated via Athena, how Lake Formation provides fine‑grained access control, and how serverless analytics (e.g., AWS Glue DataBrew, SageMaker Canvas) empower business users to build models without code.

Reference case studies include BMW’s Cloud Data Hub for global data access, a home‑appliance brand’s CDP‑driven automated marketing solution, and a major advertising client’s high‑throughput mobile marketing platform, all leveraging the intelligent data lake to break data silos and scale AI/ML workloads.

The presentation concludes that AWS helps customers evolve from data‑driven concepts to concrete implementations through integrated big‑data and AI services, unified governance, and flexible compute platforms.