AWS for Industries

Watch the re:Invent 2020 Sessions for the Advertising and Marketing Technology Industry

We were lucky to share some amazing customer stories at re:Invent 2020 focused on the advertising and marketing technology industry. With all of the announcements and keynotes, you’d be forgiven for not having time to watch each industry-specific session. That’s why we’ve compiled all of the sessions specific to advertising and marketing technology for you to watch in this post—as well as an industry playlist for you to see everything in one place. Enjoy!

Breakout sessions

How Salesforce CDP unifies consumer data at exabyte scale
Discover how Salesforce CDP enables marketers to know everything about their consumers at massive scale by using AWS. Get an inside look at how Salesforce builds on AWS to process exabytes of data, resolve known and anonymized IDs into highly accurate user profiles, and deliver audience segments across activation channels. Learn how one of the largest software companies in the world thinks about systems architecture, network topology, security, and elastic computing in order to achieve performance, cost efficiency, and reliable performance for its data platforms.

The Trade Desk — Running ultra-high throughput ad tech workloads on Aerospike
Dive deep into the AWS high-performance playbook for running ad tech workloads in the cloud using Aerospike, a low-latency NoSQL database platform. See how The Trade Desk implements Aerospike on AWS to support millions of queries per second at the edge for real-time bidding and peak loads of 30 million writes per second in its cold storage of user profiles. Learn best practices for configuring Aerospike on AWS for ad tech workloads, including recommendations for data engineering, architecture, and Amazon EC2 instance types for performance and cost efficiency.

FreeWheel, a Comcast Company – Distributed machine learning for digital video and TV ad serving
Discover how FreeWheel, a Comcast company, uses Amazon SageMaker to predict advertising inventory for digital video and linear TV months in advance for billions of ad serving records per day. Learn how FreeWheel built an end-to-end distributed ML pipeline for long-range, time-series inventory prediction across audience segments, geographies, and media types at massive scale. Additionally, take away best practices and lessons that FreeWheel learned to improve accuracy, reduce training time, optimize costs, and avoid pitfalls for ad inventory prediction with cloud-based solutions on AWS.

Cox Automotive — Building the post-cookie identity graph for marketing
See a production example of how Cox Automotive combined data from millions of car shoppers across 40,000 auto dealers and brands to create an identity graph with Amazon Neptune. Learn how to bring together customer datasets, anonymized web events, ad data, purchase logs, and other data. Take away best practices on modeling graph data, querying with Gremlin, avoiding graph pitfalls, right-sizing compute instances, and achieving cost-efficiency at scale.

Nielsen Marketing Cloud — Processing 250 billion ad events a day with Amazon EKS and serverless
Learn how Nielsen Marketing Cloud built a stateless, serverless data pipeline processing 250 billion events per day for their data management platform. See how AWS Lambda and Amazon EKS with Spark can reduce costs on logging and event processing at incredible scale, and come away with pitfalls to avoid and powers to exploit when using serverless and containers. Take away tips and tricks on how to tame the serverless beast, handle containerized processing, untangle problems delivering to real-time consumers, and avoid bugs when building your own system.

Nielsen Marketing Cloud — How Nielsen built a multi-petabyte data platform using Amazon EMR
In this session, learn how Nielsen used Amazon EMR to build and operate its multi-petabyte data lake and date warehouse. Nielsen discusses the growing pains of building a data lake, explains how to avoid them, and shares Amazon EMR best practices to improve performance in order to gain insights, reduce the cost of operating analytics workloads, and improve operational efficiency. Nielsen also walks through how it performs data exploration, sets up and shuts down Spark clusters using Jenkins, manages batch workloads through Airflow DAGs, and writes queries using notebooks.

Integral Ad Science — Contextual targeting and ad tech migration best practices
Hear how Integral Ad Science (IAS) migrated to AWS to process 100 billion events per day for its ad verification workloads—including brand safety, fraud detection, viewability measurements, and contextual advertising—supporting multiple regions and languages. IAS shares how it migrated to the cloud, overcame common industry challenges, and provided best practices on running ad tech workloads in the cloud, including architecture, data engineering, machine learning, cost analysis, migration strategy, and recommended services.

Live sessions — AWS on Air

Publicis Media — Automating Audience Segmentation with Machine Learning
Patrick Houlihan, PhD, SVP Decisioning at Publicis Media discusses how Publicis Media used Amazon EMR and SparkML to build a machine learning pipeline that trains models with petabytes of audience datasets. He shares how Publicis provides media buyers with highly accurate recommendations on audience segments. Learn how Publicis leverages machine learning to augment job roles and pairs them with natural language processing for rapid consultancy as a service.

TripleLift — Inventing Programmatic Product Placement Ads with Amazon SageMaker
TripleLift presented in “AWS on Air 2020: Industry Live Advertising and Marketing” about inventing contextually aware native ad formats for OTT and Connected TV using Amazon SageMaker, Amazon Rekognition, and AWS Elemental MediaTailor. Check out the video below to learn their architectures and see how they used machine learning with AWS. Read more about how TripleLift uses machine learning for programmatic product placement in TV advertising with AWS.

We hope you enjoyed this year’s sessions focused on the advertising and marketing industry. If you’re hungry for more re:Invent content, check out the full list of re:Invent 2020 news here or rewatch Andy Jassy’s keynote to see our biggest announcements. See you next year!

Clark Fredricksen

Clark Fredricksen

Clark Fredricksen is Head of Worldwide Marketing for the Advertising and Marketing Industry at Amazon Web Services (AWS). Before joining AWS, Clark spent a decade at research firm eMarketer, where he sat on the company’s executive management team and held leadership roles in marketing, product management, and communications. He has been quoted more than a dozen times in publications such as The New York Times, The Financial Times, The Wall Street Journal and Advertising Age speaking about trends in digital advertising and media. He is an avid cyclist.