Unlock data insights across multi-party datasets using AWS Entity Resolution on AWS Clean Rooms without sharing underlying data

The marketing and advertising technology landscape is undergoing transformation driven by fragmented consumer media and sales channels, new privacy regulations, and a rapid shift toward AI-enabled customer engagement. Companies across industries want to deliver personalized customer experiences and optimized advertising campaigns but struggle with siloed, duplicative, or missing data which hinders real-time use cases. A Gartner study reported that only 14% of companies have unified customer profiles to inform a 360-degree understanding of their customers (“Customer 360”). Even when companies have unified profiles of existing customers, the process of acquiring new customers with advertising partners (e.g., marketers and publishers, agencies, and ISVs) is challenged by the headwinds described above. The need to match and collaborate with first-party data is driving what industry experts are calling a “singularity” moment when marketing and advertising technology are merging, aided by cloud computing, modern data architectures, artificial intelligence (AI), and machine learning (ML) innovation. This merger is bringing to the forefront identity resolution and privacy-enhancing technologies (e.g., data clean rooms) as secure collaboration environments where two or more participants can analyze first-party data for mutually agreed upon uses while guaranteeing enforcement of data access limitations.

Companies are innovating and responding to these trends with the help of Amazon Web Services (AWS). AWS Entity Resolution, helps companies more easily match, link, and enhance related records across multiple applications, channels, and data stores to improve the quality of their data so that they can better understand and engage their customers. AWS Clean Rooms helps companies and their partners more easily and securely analyze and collaborate on their collective datasets—all without sharing or copying one another’s underlying data. With AWS Clean Rooms, you can create a secure data clean room in minutes and collaborate with any other company on AWS to generate unique insights about advertising campaigns, investment decisions, and research and development.

Today we announced that AWS Entity Resolution is natively integrated with AWS Clean Rooms, enabling companies to use rule-based or data service provider-based matching techniques to prepare and match their customer records with their partners’ records in privacy-enhanced AWS Clean Rooms collaborations. With a few clicks, companies and their partners can match related records, securely analyze, and collaborate on their collective datasets to build audiences, improve campaign planning, and measure campaign effectiveness —all without sharing raw data with one another.

“Protecting consumer privacy remains a top priority in the advertising and marketing industry. We are excited about AWS innovative integration of AWS Entity Resolution and AWS Clean Rooms. This integration will allow Affinity Solutions to easily link its purchase data and increase match rates with partners while protecting privacy, enabling deeper insights to drive acquisition and loyalty for our financial and retail customers.”

Logan Moore, Vice President of Business Development at Affinity Solutions

In this blog post, we discuss the benefits of using AWS Entity Resolution capabilities for AWS Clean Rooms, describe top use cases for advertising and marketing customers, and share details of how to get started with data preparation and matching to help improve data matching with collaborators.

AWS Entity Resolution on AWS Clean Rooms

Now with AWS Entity Resolution on AWS Clean Rooms, companies can use either rule-based or data service provider-based matching techniques in a collaboration. When data is matched in AWS Clean Rooms, companies benefit from privacy-enhancing rules applied to ID mapping tables, which contain data matched among collaborators’ data sets. AWS Clean Rooms protects each ID mapping table with an analysis rule that restricts members in the collaboration from directly selecting or inspecting the contents of the table.

Rule-based matching provides ready-to-use, customizable rules to prepare and match datasets among collaborators. Companies can use a no-code, rule-based matching engine powered by AWS Entity Resolution over the matching logic. Matching data by joining relevant identifiers with configurable matching rules helps save time and offers more control over the matching logic.

Data service provider-based matching provides matching with datasets and IDs from trusted data service providers—such as LiveRamp—directly within an AWS Clean Rooms collaboration. This means that companies no longer need to build ETL pipelines to generate matching outputs from a data service provider and associate them with a collaboration.

Use cases

• Media Planning: Advertisers and publishers can analyze overlapping audiences, which helps optimize advertising planning and investment.
• Audience Activation: Advertisers can collaborate with their partners to form a 360-degree view of their customers and create more accurate seed data for audience modeling with AWS Clean Rooms ML.
• Media Measurement: Measurement providers and publishers can show return on advertising spent, helping advertisers and agencies understand specific campaign outcomes. Advertisers can analyze conversion and transaction events delivered by media companies to better track and understand attribution.

Service overview

With AWS Entity Resolution on AWS Clean Rooms, companies can create data collaborations in minutes. One collaboration member can bring their first-party records to match against other collaborators’ first-party records.

For example, an advertiser may want to analyze conversion and transaction events against the impressions delivered with a publisher to analyze and understand attribution.

• The publisher, or Company A in the diagram below, creates an AWS Clean Rooms collaboration and initiates data indexing and preparation using AWS Entity Resolution to standardize data formats, remove duplicates, and ensure their dataset is ready for matching and analysis. The publisher is required to prepare their data prior to matching with their partners, unless they have already done so using AWS Entity Resolution prior to creating this collaboration.
• The advertiser, or Company B below, joins the AWS Clean Rooms collaboration and starts a data matching workflow using AWS Entity Resolution to match records with the publisher. Prior to data matching, the advertiser can optionally prepare their data using AWS Entity Resolution.

Figure 1: A data flow diagram showing the details of how to build the publisher/advertiser use case described above using AWS Entity Resolution on AWS Cleans Rooms

Rule-based matching

Rule-based entity matching in AWS Clean Rooms can be set in five steps:

1. Create or Join Collaboration: A company creates an AWS Clean Rooms collaboration and invites members to join it.
2. Create ID Namespace: From AWS Clean Rooms, each company participating in the collaboration needs to associate their customer data using AWS Entity Resolution and creating an ID namespace.
3. Associate ID Namespace: Each collaboration member will then associate their respective ID namespace to the AWS Clean Rooms collaboration.
4. Create an ID Mapping Table: A collaboration member then triggers an entity mapping workflow in AWS Clean Rooms. The output of this workflow is an ID mapping table which becomes available in AWS Clean Rooms.
5. Run Queries in AWS Clean Rooms: Based on agreed-upon collaboration constraints among all members, a participant can run queries in AWS Clean Rooms for various use cases, such as planning and measurement. The ID mapping table can be queried in conjunction with configured tables including any analysis rule type (list, aggregation, or custom SQL) added by collaboration members.

Figure 2: An architecture diagram illustrating the five-step workflow when rule-based matching is selected.

Data service provider-based matching

Data service provider-based matching in AWS Clean Rooms can be set in five steps:

1. Create or Join Collaboration: A company creates an AWS Clean Rooms collaboration and invites members to join it.
2. Create ID Namespace: From AWS Clean Rooms, each company participating in the collaboration needs to associate their customer data using AWS Entity Resolution. In AWS Entity Resolution, companies create an ID namespace that helps them specify the set of RampIDs they want to transcode. An ID namespace is a resource that each customer in the collaboration creates to refer to their set of RampIDs and choose a role for the direction of transcoding. The customer that wishes to initiate the transcoding selects “source”. The other collaborator selects “target”.
3. Associate an ID Namespace: Each collaboration member will then associate their respective ID namespace to the AWS Clean Rooms collaboration.
4. Create an ID Mapping Table: The collaboration member initiating the transcoding (for example, the advertiser) creates an ID mapping table.
5. Run Queries in AWS Clean Rooms: Using the ID mapping table, collaboration members can run queries in AWS Clean Rooms for various use cases, such as campaign planning and measurement. This ID mapping table is used in a collaboration with a set of pre-defined, immutable analysis rules to restrict analysis to only SQL JOIN statements. Additionally, companies can use the ID mapping table in conjunction with AWS Clean Rooms ML to accept a SQL query as the seed data source for lookalike audience modeling.

Figure 3: An architecture diagram of the five-step workflow when data service provider-based matching is selected.

Conclusion

In this post, we have showed how companies can easily match their data with their collaborators’ datasets using rule-based and data service provider-based matching techniques for use cases including advertising campaign planning, lookalike modeling, measurement—all without sharing their underlying data with each other.

If you’d like to learn more about AWS Entity Resolution on AWS Clean Rooms check out our website or contact a privacy-enhanced data collaboration expert.

Additional resources

AWS Entity Resolution on AWS Clean Rooms pricing

AWS Clean Rooms User Guide

AWS Entity Resolution website