AWS Security ChangesHomeSearch

AWS wellarchitected: VSA Lens: add cross-cutting design principles incl. privacy/security

Service: wellarchitected · 2026-11-11 · Documentation medium

File: wellarchitected/latest/video-streaming-advertising-lens/design-principles.md

Summary

Restructures the Video Streaming Advertising Lens design-principles page into lens-wide "Cross-cutting design principles" plus a "Pillar-level design principles" index linking to each pillar. The prior single list of principles is replaced by expanded narratives, including a new principle "Design for privacy, security, and regulatory adherence" that calls for strong access control, encryption, fraud detection, privacy-enhancing technologies (data clean rooms, confidential computing), consent/data-rights management, and alignment to GDPR/CCPA and programmatic advertising standards. Boilerplate footer text is also updated.

Security assessment

The added line explicitly prescribes security and privacy design guidance — treating partners as independent trust domains with access control and encryption, plus fraud detection, PETs, and consent/regulatory handling — which is general best-practice documentation rather than a fix for a specific vulnerability or incident, so it is security documentation but not issue-related.

Evidence

+  * **Design for privacy, security, and regulatory adherence:** Treat every partner as an independent trust domain with strong access control, encryption, and industry-standard fraud detection. Address third-party signal loss with privacy-enhancing technologies such as data clean rooms and confidential computing, first-party data, and contextual targeting. Manage consent and data rights, and treat regional and industry regulations (GDPR, CCPA, and programmatic advertising standards) as a design constraint rather than an afterthought. 

Diff

diff --git a/wellarchitected/latest/video-streaming-advertising-lens/design-principles.md b/wellarchitected/latest/video-streaming-advertising-lens/design-principles.md
index 0fdb132..f193d79 100644
--- a//wellarchitected/latest/video-streaming-advertising-lens/design-principles.md
+++ b//wellarchitected/latest/video-streaming-advertising-lens/design-principles.md
@@ -6,0 +7,2 @@
+Cross-cutting design principlesPillar-level design principles
+
@@ -9 +11,22 @@
-The following design principles can help you achieve and maintain efficient advertising workloads in the cloud. 
+These design principles underpin all six pillars of the Video Streaming Advertising (VSA) Lens. They describe how to think about designing, building, and operating advertising workloads in the cloud. Individual pillar-level design principles extend these lens-wide principles with domain-specific guidance. 
+
+## Cross-cutting design principles
+
+  * **Design for low latency and elastic scaling:** Advertising traffic swings by an order of magnitude with live events, campaign launches, and seasonal peaks, and much of it runs under real-time bidding deadlines. Build horizontally scalable, containerized services behind load balancers, drive auto scaling from application metrics, and cache hot data and content to keep latency low and reduce database load. Choose memory- and CPU-optimized hardware where ultra-low-latency bid paths require it. 
+
+  * **Design for high-volume data and scalable pipelines:** Advertising generates large volumes of bid, impression, and event data at high write rates across regions. Favor distributed, write-optimized data stores, streaming ingestion, and event-driven precomputation over synchronous processing on the critical path. Combine real-time streaming with batch analytics, aggregate into a cataloged data lake, and apply compression, tiered storage, and archival policies to control cost, while monitoring latency, timeouts, and saturation to find bottlenecks. 
+
+  * **Design for cost efficiency at advertising scale:** The main cost drivers are data transfer, compute, storage, and networking, and at campaign volumes even small per-request savings compound. Minimize wasted traffic and processing, and provide sufficient but not excessive capacity through auto scaling rather than static over-provisioning. 
+
+  * **Design for privacy, security, and regulatory adherence:** Treat every partner as an independent trust domain with strong access control, encryption, and industry-standard fraud detection. Address third-party signal loss with privacy-enhancing technologies such as data clean rooms and confidential computing, first-party data, and contextual targeting. Manage consent and data rights, and treat regional and industry regulations (GDPR, CCPA, and programmatic advertising standards) as a design constraint rather than an afterthought. 
+
+  * **Design for generative and agentic AI:** Use generative AI for creative optimization, contextual targeting, and personalized experiences at scale, and agentic workflows to automate decisions across campaign management, bid optimization, and audience segmentation. Adopt standardized industry frameworks (AdCP and ARTF from IAB Tech Lab), and apply responsible-AI practices with human oversight, bias detection, and explainability. 
+
+  * **Design publisher monetization for automation and linear-digital convergence:** Automate the chain from content through ad insertion to revenue reporting, and keep auction, playout, and channel-assembly components stateless and horizontally scalable so they absorb traffic spikes without manual intervention. Converge linear and digital operations onto shared infrastructure and a unified data foundation for campaign planning, fulfillment, pacing, and cross-platform measurement. 
+
+
+
+
+## Pillar-level design principles
+
+For design principles specific to each Well-Architected pillar, see: 
@@ -11 +34 @@ The following design principles can help you achieve and maintain efficient adve
-  * **Design for low latency and rapid changes in traffic volume:** Build a scalable architecture with automated scaling capacity to enable rapid increases and decreases of traffic gracefully. Use load balancers to distribute traffic to new nodes available to scale horizontally. Implement auto scaling based on application node metrics. Cache application data content, when possible, to reduce response latency and the load on database clusters. Use containerized workloads and prebuilt container images for fast scaling and predictable performance. Choose server hardware optimized for memory and CPU for ultra-low latency needs. 
+  * [Operational Excellence design principles](./ops-design-principles.html)
@@ -13 +36 @@ The following design principles can help you achieve and maintain efficient adve
-  * **Design for rapid growth, very large data volumes, high QPS (queries per second), and low latency transactions across multiple Regions:** Build a scalable distributed database for transactions while optimizing it for fast writes. Consider use of a distributed NoSQL database that can handle high write throughput with linear scalability. Consider compression techniques to optimize storage and an appropriate caching strategy to reduce database load for user profiles, target segments, and creatives. Use streaming services for ingestion and transportation of event data. Set up auto scaling databases to handle traffic spikes. Implement a data archive strategy to purge old ad impressions data to less expensive storage. Monitor database performance metrics, including latency, timeouts, and saturation, to identify and fix bottlenecks. 
+  * [Security design principles](./sec-design-principles.html)
@@ -15 +38 @@ The following design principles can help you achieve and maintain efficient adve
-  * **Design for cost optimization to reduce costs while maintaining performance:** The key to cost optimization for advertising workloads is to minimize costs while you maintain performance and reliability. Optimization efforts should focus on minimizing unnecessary traffic charges and providing sufficient but not excessive capacity through auto scaling. The main drivers of advertisement costs are data transfer, compute, storage, and networking. Considering the large amounts of traffic involved in campaigns, even small changes can yield significant cost optimization. 
+  * [Reliability design principles](./rel-design-principles.html)
@@ -17 +40 @@ The following design principles can help you achieve and maintain efficient adve
-  * **Design for scaled, cost-optimized, and performant data pipelines:** Use event-driven architecture and pre-compute data where possible to reduce real-time compute needs. Design for horizontal scalability rather than optimizing within a single host. For large-scale systems, consider distributed service-oriented architecture to allow components to innovate independently. Use caching solutions at scale, with multiple replicas for high availability and data recovery. Implement real-time introspection and debugging capabilities for system transparency. When dealing with massive data volumes, consider batching and direct writing to object storage instead of using intermediary services. Use a combination of real-time streaming and batch analytics processing for data processing, based on specific use cases. Implement a centralized data lake for aggregating data from various sources with a data catalog. Use tiered storage solutions to balance cost and performance for data storage. Reuse compute resources like analytics clusters, to reduce costs and startup times. Retain all event-level data and logs for less than 30 days. 
+  * [Performance Efficiency design principles](./perf-design-principles.html)
@@ -19 +42 @@ The following design principles can help you achieve and maintain efficient adve
-  * **Design for privacy-enhanced collaboration:** Consider the following strategies to address third-party signal loss through linked first-party data, browser and OS-mediated data, and unlinked first-party data. Use linked first-party data design, where advertiser and publisher data are connected through privacy-enhancing technologies (PETs), enabling personalized customer engagement and targeting while maintaining privacy. Use browser and OS-mediated data design where web browsers or operating systems act as intermediaries for user data collection and sharing. It protects user privacy by limiting direct access to personal information, instead of providing aggregated or anonymized data to advertisers and websites. Use contextual targeting to deliver relevant ads without relying on user data and seller-defined audience solutions which allow publishers to create targetable segments based on their first-party data. Implement first-party and third-party identity matching and identity transcoding that use trusted and compliant runtime environments like data clean rooms and confidential computing environments. Design for robust access controls, audit trails, and secured data exchange mechanisms for your organization and partners. Design for alignment on file formats, APIs, communication protocols, and data schemas, ideally with standardized industry formats as defined by the Interactive Advertising Bureau (IAB). 
+  * [Cost Optimization design principles](./cost-design-principles.html)
@@ -21 +44 @@ The following design principles can help you achieve and maintain efficient adve
-  * **Design for advertising regulatory, security, and privacy needs:** Design for robust access control and network security measures, complemented by industry-standard ad fraud detection software and anti-fraud tools. Design for monitoring and assessment of ads to facilitate policy compliance and malicious content detection, while maintaining strong consumer privacy protection, and managing consent and data rights. Consider regulatory standards involved for ad formats and creative standards (IAB), measurement standards (MRC, IAB), privacy and consent standards (IAB, regional regulations for example, China, CCPA, and GDPR), programmatic advertising standards (IAB, openRTB consortium), brand safety and fraud prevention (TAG, IAB), identity and attribution (U-ID2.0, rampID, ID5, MTA) and specific ad standards published by Google, Meta, and Amazon. 
+  * [Sustainability design principles](./sus-design-principles.html)
@@ -32 +55 @@ To use the Amazon Web Services Documentation, Javascript must be enabled. Please
-List of AWS services 
+Definitions