AWS wellarchitected: Rewrite ADVOPS01-BP02 RACI guidance, remove security content
Summary
Rewrites the RACI best practice to focus on publisher-specific operational roles and removes prior data governance content including IAM least privilege, SCPs, Control Tower, and AWS Config compliance guidance.
Security assessment
The diff is a documentation rewrite of a Well-Architected best practice. It removes security-adjacent data governance guidance (IAM least privilege, SCPs, Control Tower, AWS Config) and replaces it with operational RACI content. No specific vulnerability, CVE, or incident is referenced, and no new security feature is documented.
Evidence
- * Configure IAM roles with least-privilege permissions aligned to team responsibilities (for example, real-time team with write access to bidding data, read-only for analytics)
Diff
diff --git a/wellarchitected/latest/video-streaming-advertising-lens/advops01-bp02.md b/wellarchitected/latest/video-streaming-advertising-lens/advops01-bp02.md index ae924d0..c393cea 100644 --- a//wellarchitected/latest/video-streaming-advertising-lens/advops01-bp02.md +++ b//wellarchitected/latest/video-streaming-advertising-lens/advops01-bp02.md @@ -7 +7 @@ -Implementation guidanceResourcesKey AWS services +Implementation guidanceResources @@ -9 +9 @@ Implementation guidanceResourcesKey AWS services -# ADVOPS01-BP02 Create RACI matrices that define the roles and responsibilities for each key advertising process like infrastructure monitoring +# ADVOPS01-BP02 Create RACI matrices that define the roles and responsibilities for each key advertising process @@ -11 +11 @@ Implementation guidanceResourcesKey AWS services -When designing advertising workloads, define roles and set clear expectations for each stakeholder for seamless key advertising processes. By implementing this best practice, organizations can leverage RACI (responsible, accountable, consulted, and informed) matrices to establish a robust framework for accountability and decision-making. +Define stakeholder roles so key advertising processes run without handoff gaps. Use RACI (responsible, accountable, consulted, and informed) matrices to assign accountability and decisions for publisher-specific workloads (SGAI manifest pipelines, prebid adapter onboarding, and FAST channel scheduling) and for converged linear and digital operations. @@ -13 +13 @@ When designing advertising workloads, define roles and set clear expectations fo -## Implementation guidance +**Desired outcome:** @@ -15 +15 @@ When designing advertising workloads, define roles and set clear expectations fo -By creating comprehensive RACI matrices, organizations can establish accountability, decision-making authority, and communication requirements for each step of the ad-serving workflow. This level of clarity assists in blocking confusion, gaps, or overlaps in responsibilities. This clarity also verifies that stakeholders understand their roles and how they contribute to the overall success of the advertising operations. + * Publisher-specific RACI matrices clearly define accountability for Prebid adapter lifecycle management, SGAI manifest generation pipeline ownership, FAST channel scheduling and playout monitoring, and innovative ad format rendering operations. @@ -17 +17 @@ By creating comprehensive RACI matrices, organizations can establish accountabil -For data management specifically, implement the following approach: + * Each publisher theme has a designated operational owner accountable for yield outcomes, not just system health. @@ -19 +19 @@ For data management specifically, implement the following approach: - 1. Establish data classification and handling processes: + * For converged environments, a unified command structure bridges the linear broadcast NOC and digital ad operations with pre-defined cross-functional escalation paths. @@ -21 +20,0 @@ For data management specifically, implement the following approach: - * Classify advertising data based on criticality and latency requirements (real-time bidding data, campaign configuration data, historical analytics data) @@ -23 +21,0 @@ For data management specifically, implement the following approach: - * Define data retention policies for each classification (for example, bid data retained for 30 days, campaign data for one year) @@ -25 +22,0 @@ For data management specifically, implement the following approach: - * Implement data lineage tracking using AWS Glue Data Catalog to document data origins, transformations, and dependencies across the advertising pipeline @@ -27 +24 @@ For data management specifically, implement the following approach: - 2. Structure teams around data criticality: +**Common anti-patterns:** @@ -29 +26 @@ For data management specifically, implement the following approach: - * **Real-time data operations team:** Responsible for sub-100ms data like bidding, user profiles, and fraud detection + * Assigning SGAI manifest generation ownership to infrastructure teams without ad operations input on SCTE-35 marker configuration, resulting in technically correct manifests that miss ad break opportunities. @@ -31 +28 @@ For data management specifically, implement the following approach: - * **Campaign data management team:** Handles near real-time data for configurations, targeting, and budgets + * Leaving Prebid adapter onboarding accountability undefined between engineering and partner management teams, causing delays when new demand partners need to be integrated. @@ -33 +30 @@ For data management specifically, implement the following approach: - * **Analytics and reporting team:** Manages batch processing for historical data and business intelligence + * Operating FAST channels without clear ownership boundaries between content scheduling, playout engineering, and ad operations functions. @@ -35 +32 @@ For data management specifically, implement the following approach: - 3. Define data ownership with specific domains: + * Maintaining separate RACI matrices for linear and digital teams, with no cross-functional ownership defined for converged failure scenarios. @@ -37 +33,0 @@ For data management specifically, implement the following approach: - * Assign data stewards for specific advertising domains such as: @@ -39 +34,0 @@ For data management specifically, implement the following approach: - * Bid management domain (bid requests, responses, auction data) @@ -41 +35,0 @@ For data management specifically, implement the following approach: - * User profile domain (demographic data, behavioral signals, privacy preferences) @@ -43 +37 @@ For data management specifically, implement the following approach: - * Campaign domain (creative assets, targeting parameters, budget configurations) +**Benefits of establishing this best practice:** @@ -45 +39 @@ For data management specifically, implement the following approach: - * Analytics domain (performance metrics, attribution data, reporting datasets) + * Clear accountability reduces coordination delays during incidents. @@ -47 +41 @@ For data management specifically, implement the following approach: - * Document domain-specific quality standards and governance responsibilities in the RACI matrix + * Faster decision-making through pre-defined escalation paths. @@ -49 +43 @@ For data management specifically, implement the following approach: - 4. Implement data governance using AWS services: + * Reduced ownership gaps where issues that affect revenue fall between team boundaries. @@ -51 +44,0 @@ For data management specifically, implement the following approach: - * Use AWS Organizations with service control policies (SCPs) to enforce data residency requirements for different Regions @@ -53 +45,0 @@ For data management specifically, implement the following approach: - * Configure IAM roles with least-privilege permissions aligned to team responsibilities (for example, real-time team with write access to bidding data, read-only for analytics) @@ -55 +46,0 @@ For data management specifically, implement the following approach: - * Deploy AWS Control Tower guardrails to help block unauthorized cross-account or cross-Region data transfers @@ -57 +48 @@ For data management specifically, implement the following approach: - * Implement AWS Config rules to continuously audit compliance with data governance policies +**Level of risk exposed if this best practice is not established:** Medium @@ -59 +50 @@ For data management specifically, implement the following approach: - 5. Establish data management processes: +## Implementation guidance @@ -61 +52 @@ For data management specifically, implement the following approach: - * Data cataloging: Use AWS AWS Glue Data Catalog to maintain a comprehensive inventory of advertising datasets with metadata, ownership, and classification +Engineering owns infrastructure, ad operations owns yield outcomes, partner management owns DSP and SSP relationships, and content teams own programming schedules. Without explicit accountability mapping, cross-boundary incidents create coordination overhead. @@ -63 +54 @@ For data management specifically, implement the following approach: - * Quality monitoring: Implement automated data quality checks using AWS Glue DataBrew to validate incoming data against defined schemas and business rules +For SGAI, the manifest generation pipeline exists at the intersection of media infrastructure (transcoding, packaging, and CDN) and advertising (ad break timing, SCTE-35 accuracy, and yield reporting). Neither team can own it independently, so the RACI must assign responsible parties from both domains with explicit coordination points. For Prebid Server, adapter onboarding involves engineering (container deployment and configuration), partner management (commercial agreements and technical integration specs), and ad operations (yield monitoring after launch). Without a documented process owner, adapters accumulate unmaintained. @@ -65 +56 @@ For data management specifically, implement the following approach: - * Workflow automation: Create AWS Step Functions workflows for data handoffs between teams, with validation checkpoints and approval gates for critical data transitions +Converged linear and digital operations need a Tier 2 cross-functional incident commander with authority to bridge the linear NOC and digital ad ops teams during incidents. This person needs pre-approved authority to escalate, request resources, and make trade-off decisions. The RACI must define this role and identify who holds it per shift. @@ -66,0 +58 @@ For data management specifically, implement the following approach: +Assign data stewards to specific advertising domains, such as bid management, user profile, campaign, and analytics. Each steward is accountable for the quality, retention, and access controls of their domain's data. @@ -67,0 +60 @@ For data management specifically, implement the following approach: +### Implementation steps @@ -68,0 +62 @@ For data management specifically, implement the following approach: + 1. **Identify all publisher-specific operational processes requiring RACI definition:** Catalog processes across adapter onboarding, manifest pipeline management, FAST channel scheduling, ad format rendering, and converged linear and digital campaign management. For each process, identify the teams involved: engineering, ad operations, partner management, content, and any external partners. @@ -70 +64 @@ For data management specifically, implement the following approach: -## Resources + 2. **Map each process to RACI roles:** For each identified process, define responsible (who does the work), accountable (who is answerable for the outcome), consulted (who provides input), and informed (who is notified of outcomes). For publisher yield processes, the accountable role must sit with whoever owns the revenue outcome, typically ad operations for processes that affect yield. @@ -72 +66 @@ For data management specifically, implement the following approach: - * AWS Organizations for team boundaries: Implement multi-account strategies with SCPs to enforce separation of duties between data teams as described in [AWS Organizations Best Practices](https://docs.aws.amazon.com/organizations/latest/userguide/orgs_best-practices.html) + 3. **Define handoff points and escalation triggers at team boundaries:** Document the specific conditions that trigger escalation from one team to another. For SGAI: if manifest error rates exceed X%, escalate from infrastructure to ad operations and initiate the joint runbook. For prebid: if a new adapter's win rate is below Y% after 48 hours, escalate from ad operations to engineering for technical review. Clear escalation criteria help teams avoid debating ownership during an incident. @@ -74 +68 @@ For data management specifically, implement the following approach: - * Data governance implementation: Follow the framework in [AWS Data Governance Whitepaper](https://docs.aws.amazon.com/whitepapers/latest/data-governance-aws/) to establish controls specific to advertising data domains + 4. **Define the cross-functional incident commander role for converged operations:** For organizations running both linear broadcast and digital ad serving, document a Tier 2 incident commander role with a 15-minute escalation authority. Define the conditions that activate this role (incidents affecting both linear and digital delivery simultaneously), the pre-approved actions this role can take without further approval, and the on-call rotation that covers this role. @@ -76 +70 @@ For data management specifically, implement the following approach: - * Team collaboration on data assets: Use [Amazon DataZone](https://docs.aws.amazon.com/datazone/latest/userguide/) to create a data marketplace where teams can discover, share, and collaborate on advertising datasets + 5. **Publish RACI matrices and review quarterly:** Store RACI matrices in a shared operational wiki accessible to all relevant teams. Schedule quarterly reviews to update accountability when team structures change, when new workload types are onboarded, and after incident retrospectives that reveal ownership gaps. Update the RACI before changes go live, so an incident isn't the first time a team discovers an ownership gap. @@ -78 +71,0 @@ For data management specifically, implement the following approach: - * Automated operational procedures: Implement [AWS Systems Manager](https://docs.aws.amazon.com/systems-manager/latest/userguide/systems-manager-best-practices.html) runbooks for standardized data operations tasks across teams @@ -80 +72,0 @@ For data management specifically, implement the following approach: - * Compliance monitoring: Deploy [AWS Config rules](https://docs.aws.amazon.com/config/latest/developerguide/best-practices.html) to continuously validate that data handling practices meet organizational and regulatory requirements @@ -82 +73,0 @@ For data management specifically, implement the following approach: - * [Create a RACI or RASCI matrix for a cloud operating model](https://docs.aws.amazon.com/prescriptive-guidance/latest/patterns/create-a-raci-or-rasci-matrix-for-a-cloud-operating-model.html) @@ -83,0 +75 @@ For data management specifically, implement the following approach: +## Resources @@ -84,0 +77 @@ For data management specifically, implement the following approach: +**Related best practices:** @@ -85,0 +79 @@ For data management specifically, implement the following approach: + * [ADVOPS01-BP03 Establish performance metrics by defining key performance indicators (KPIs) and service-level objectives (SLOs)](./advops01-bp03.html) @@ -87 +81 @@ For data management specifically, implement the following approach: -## Key AWS services + * [ADVOPS04-BP01 Create runbooks for the most common operational events and incidents that can impact your advertising workload](./advops04-bp01.html) @@ -89 +82,0 @@ For data management specifically, implement the following approach: - * AWS Organizations @@ -91 +83,0 @@ For data management specifically, implement the following approach: - * AWS IAM @@ -93 +84,0 @@ For data management specifically, implement the following approach: - * AWS Control Tower @@ -95 +86,6 @@ For data management specifically, implement the following approach: - * AWS AWS Glue Data Catalog +**Related documents:** + + * [Create a RACI or RASCI matrix for a cloud operating model](https://docs.aws.amazon.com/prescriptive-guidance/latest/patterns/create-a-raci-or-rasci-matrix-for-a-cloud-operating-model.html) + + * [AWS Organizations best practices](https://docs.aws.amazon.com/organizations/latest/userguide/orgs_best-practices.html) + @@ -97 +92,0 @@ For data management specifically, implement the following approach: - * AWS Glue DataBrew @@ -99 +93,0 @@ For data management specifically, implement the following approach: - * Amazon DataZone @@ -101 +95 @@ For data management specifically, implement the following approach: - * AWS Systems Manager +**Related services:** @@ -103 +97 @@ For data management specifically, implement the following approach: - * AWS Config + * [AWS Organizations](https://aws.amazon.com/organizations/) @@ -105 +99 @@ For data management specifically, implement the following approach: - * AWS Step Functions + * [AWS IAM](https://aws.amazon.com/iam/)