AWS for M&E Blog

AWS demonstrates new monetization solutions at IBC 2023

With the announcement of the AWS for M&E Monetization solution area, Amazon Web Services (AWS) demonstrates new monetization solutions at IBC 2023 that enable media companies to maximize advertising revenue and drive customer engagement across platforms and channels.

Ten AWS and AWS Partners feature solutions spanning the end-to-end media advertising lifecycle on stand 5.C90 at the RAI Exhibition Centre in Amsterdam September 15 – 18. Demonstrations include customer 360 (C360) solutions focused on attracting and retaining audiences, along with converged TV solutions to unify linear and digital campaign planning and execution, optimize ad inventory across channels, and enable data collaboration for unified audience measurement. Visitors to the stand can also view demos of advanced advertising innovations such as contextual advertising, and new, non-intrusive video ad formats such as virtual product placement and shoppable video. Use cases for generative AI in advertising are featured, highlighting ad creative generation targeted to specific audiences or individual consumers.

Customer 360 Demonstrations

Composable C360 with Databricks and Hightouch

Composable C360 solutions allow customers to build enriched views of their consumers by activating data sets within their own data warehouse, instead of sending that data to an external customer data platform (CDP). The modular nature of the architecture lets customers stitch together best-of-breed components as an extension of their broader customer data management strategy. Self-service, modular platforms can be created to generate audience insights, customer attraction and retention models, and customer journeys.

At IBC, AWS features a demo that uses the Databricks Lakehouse platform as the storage and data modeling layer of the composable C360. Raw subscriber, click-stream, and quality of service (QoS) data is loaded into the lakehouse using Delta Live Tables. The C360 view is built by enriching, merging, and aggregating these raw data tables. Machine learning platforms such as MLflow and AutoML natively integrate into Databricks, letting data scientists easily train and deploy customer engagement models, such as propensity to churn as in this demo. The Hightouch reverse ETL data activation platform streams churn prediction data from the lakehouse for use in any business application.

AWS Services – Amazon S3
Partners – DataBricks, HighTouch

Hybrid-Composable Customer 360 with ActionIQ

A hybrid composable C360 solution lets customers sync customer data directly from their own data warehouses, without porting the data to a CDP. In this scenario, data residency stays fully within the customer environment, enabling marketing and advertisers to create C360 solutions to define audiences, create journeys, and stand up campaigns in a self-service environment.

The IBC demo uses ActionIQ’s HybridCompute (InfiniteCompute) connector, integrated with an Amazon Redshift data warehouse to natively support a hybrid-composable C360 deployment, to provide full control over where data is accessed and queried. The HybridCompute connector allows for query virtualization and post processing, making cross-platform query results accessible by pushing low/no-code queries directly to the data lake. Results can be activated directly to ActionIQ’s CX Hub or third-party business applications.

AWS Services – Amazon S3, AWS Glue, Amazon Redshift
Partners – ActionIQ

Converged TV Advertising Demonstrations

Unified Ad Sales on AWS

AWS addresses the need to simplify and unify the linear and digital advertising ecosystem by facilitating cross-platform ad buying, fulfillment, and measurement. These solutions optimize and automate the ad sales process, maximizing yield across direct and programmatic sales, and ensuring campaigns are fulfilled across platforms with minimum waste.

The IBC demo features a single pane of glass view for campaign management, unified measurement, and data-driven inventory placement and optimization. Unified sales planning tools facilitate selling of inventory across all platforms and devices from one place, while best-in-class data science-driven algorithms ensure optimal placement of inventory across channels. Unified measurement tools enable comparability, coverage, and consistency of multiple currencies, leveraging data sets from multiple measurement providers.

AWS Services: AWS Direct Connect, VPN, Amazon Route 53, AWS WAF, Amazon S3, Amazon API Gateway, Amazon ECR, AWS Fargate, Amazon ECS, Amazon Aurora, Amazon DynamoDB, Lake Formation, AWS Glue, Amazon AppFlow, Amazon API Gateway, Amazon Kinesis, AWS DataSync, AWS DMS, Amazon SageMaker, Amazon EMR, Amazon Athena, Amazon QuickSight, Amazon EventBridge, Amazon CloudTrail, Amazon CloudWatch, AWS CodePipeline, AWS Security Hub

Data Collaboration for Unified Measurement and Audience Activation on AWS Clean Rooms

Data collaboration for audience enrichment, measurement, and activation is enabled via a drag-and-drop UI with a no-code interface on AWS Clean Rooms. Business decision makers can instantiate multi-party data analyses without using the AWS Console or writing a single line of code. An abstraction layer on top of AWS Clean Rooms service enables no-code operations for selecting data sets for analysis, merging tables, defining groupings and aggregations, applying filters, and publishing results to an analytics suite.

The IBC demo showcases ease of collaboration on AWS Clean Rooms with enriched first party predictive household data from ThinkAnalytics. This collaboration can also be extended to enriched, cross-platform audience measurement via data sets from providers like iSpot and Comscore.

AWS Services – AWS Clean Rooms, AWS Amplify, Amazon CloudFront, Amazon Cognito, Amazon DynamoDB, AWS Lambda, Amazon S3, Amazon CDK, AWS CloudFormation
Partners – ThinkAnalytics, iSpot, Comscore

Advanced Advertising Demonstrations

Advanced Advertising Formats with AWS Elemental MediaTailor

At IBC, AWS features new, audience friendly video advertising formats supported by AWS Elemental MediaTailor. The booth demo features non-intrusive ad formats – squeeze backs/frame ads, ad overlays, and in-scene virtual product placement. These advanced ad formats appear at the same time as the primary content, allowing continuity of viewership, as opposed to existing pre-roll and mid-roll ad formats that take the viewer away from their content. Overlay advertising can be served across a range of devices, including those without native overlay ad support (based on the device’s capability).

AWS Services – AWS Elemental MediaTailor SSAI, AWS Elemental Channel Assembly
Partners – Innovid, TripleLift, Castlabs, SpringServe

Just-In-Time Ad Pod Construction 

Deferring ad pod construction to immediately before an ad pod is played results in faster video startup times, more efficient interactions with ad tech partners, higher fill rates, higher render rates, and better revenue performance for ad supported video on demand (AVOD) content. Server-side ad insertion (SSAI) with client-side telemetry provides a superior viewer experience than client-side ad insertion (CSAI), with higher ad revenue potential.

At IBC, AWS Elemental MediaTailor SSAI, with Penthera’s 2nd Look, and Datazoom demonstrate a smooth, low-latency viewer experience with seamless transition between content and ads, no ad or content failures, consistent quality and audio levels across content and ads, all in a cloud-based solution with no software to manage and low susceptibility to ad blockers.

AWS Services – AWS Elemental MediaTailor SSAI, Amazon S3, Amazon CloudFront
Partners – Penthera, Datazoom

Frame-Accurate Ad Break Detection and Contextual Advertising

At IBC, AWS and Partners demonstrate a solution that identifies ad break opportunities between scene changes, passes contextual metadata for ad decisioning, and selects ads that are contextually relevant to the content. Selection criteria are a combination of scene changes, dialogue, silence, and custom business logic. Candidate ads are given relevant weights and confidence scores. Contextual metadata pre- and post-ad break opportunities are passed to the ad proxy service or ad decisioning server (ADS) for matching the context of content to ads, and prioritizing the serving of contextually relevant ads.

AWS ServicesAmazon Rekognition, Amazon Transcribe, Amazon EventBridge, AWS Lambda, AWS Elemental MediaConvert, AWS Elemental MediaPackage, Amazon CloudFront, Amazon API Gateway, Amazon DynamoDB, Amazon S3
AWS Solutions – Media2Cloud

Shoppable Video

Shoppable video is an immersive experience for viewers that provides a seamless in-content shopping experience. When a user presses pause during programming, an “X-Ray” like experience is initiated, showing shoppable items within the lower-third of the paused scene.

At IBC, AWS demonstrates this experience, with a solution that simplifies identification of shoppable images (such as apparel) by automating the human metadata labeling effort. Once shoppable items are identified, image embeddings (numeric representations of shoppable items) are matched to content embeddings in an e-commerce vector store. Based on a similarity score, similar items are represented in a shopping store and presented to the viewer with a checkout experience without leaving their viewing content. As a serverless solution, it can be integrated with front end UI systems via API, providing flexibility to customize the viewer experience.

AWS Services – Amazon SageMaker, Amazon Rekognition, Amazon OpenSearch, AWS Step Functions, AWS Lambda, Amazon DynamoDB, AWS Elemental MediaConvert, Amazon API Gateway, Amazon S3

Generative AI for Advertising Demonstrations

Ad Creative Generation with Stable Diffusion XL 

Generative AI can be used to generate images for ad campaigns, such as base images for products tailored to specific audiences. Using a Stable Diffusion large language model (LLM) to generate ad creative images, and leveraging first party data from a CDP, individual, consumer profile-specific images can be generated.

AT IBC, AWS demonstrates a solution using Stability AI on Amazon SageMaker Jumpstart to create base image with first party profile data. The solution incorporates prompts including demographic attributes such as gender, nationality, lifestyle, age, and income, which are fine-tuned to create hyper-personalized, specialized images. Various types of image styles can be created, including photographic, anime, digital art and cinematic.

AWS Services – Amazon SageMaker, Amazon Amplify, Amazon Cognito, S3, Amazon API Gateway, AWS AppSync, AWS Lambda, Amazon DynamoDB, Amazon SQS, Amazon ECS, Amazon CDK, AWS CloudFormation

Ad Creative Generation with Stable Diffusion XL from Stability AI and Ad Tag Line Generation with Claude V2 from Anthropic on Amazon Bedrock

Stable Diffusion and Claude LLM models are used to generate ad creative images and ad tag lines, in a fully managed, serverless infrastructure.

The IBC demo uses generative AI to generate ad creative images from text-to-image or image-to-image prompts. Text-to-image and image-to-image generation allows for positive and negative prompts, styling, clip guidance presets, samplers, cfg scale, seeds, and steps. Ad tag line generation allows for free-form product description prompts, gender, and target age groups.

AWS Services: Amazon Bedrock, Amazon CloudFront, Amazon S3, Amazon Cognito, Amazon API Gateway, AWS Lambda, Amazon DynamoDB

Schedule a meeting to see AWS and AWS Partner demonstrations at IBC 2023 and learn how AWS is helping media companies modernize their monetization ecosystems. Find AWS on stand 5.C90.

Michael Santana

Michael Santana

Michael Santana is a Sr Specialist Solutions Architect, AdTech/MarTech and the technical lead for the newly launched Monetization solution area for AWS M&E. Michael led the development efforts on the Monetization launch demos previewed at IBC 2023. By training, Michael is an applied physicist and pure mathematician. Since having started his career at Goldman Sachs as a quantitative strategist, he has held senior leadership roles in engineering and data science at Roku, Disney and Paramount.

Chris Ziemer

Chris Ziemer

Chris Ziemer is a Media & Entertainment Industry Specialist at AWS. Chris works with M&E business and technology leaders to accelerate cloud transformation initiatives, providing domain expertise for cloud adoption in broadcast, ad-tech transformation, live production, media supply chain, and streaming media workflows. Prior to AWS, Chris held senior sales, business development, solutions engineering, and alliance partner management positions at Imagine Communications and Harris Broadcast. Chris has led some of the industry’s most transformative technology projects with global media companies and broadcasting groups.