Skip to main content

Getting Started with LiveRamp Clean Room for Publishers

LiveRamp Clean Room offers a streamlined, one-time integration that connects publishers to their data assets, enabling collaboration across the ecosystem:

  • Interoperable: Collaborate in one place with your brand partners globally, no matter the cloud

  • Flexible: Supports code-based analytics for technical users and no-code UX for business users to generate insights and reporting

  • Actionable: Turn insights into action with identity-powered activation to 350+ destinations

LiveRamp Clean Room Overview

A clean room is a secure environment that enables data collaboration across decentralized datasets to help uncover actionable insights and outcomes.

image idm649

LiveRamp Clean Room enables the creation and ongoing support of clean room collaborations through a user-friendly interface.

Included in LiveRamp Clean Room's architecture are three core layers:

image idm654
  • The Integration Layer, where we integrate with cloud and media ecosystems to access data at its source. This layer provides foundational support for use cases ranging from reporting insights to user list creation, complex modeling, and more.

  • The Application Layer, which enables teams to automate dataset workflows and define and enforce governance rules and privacy controls for multi-party collaborations.

  • The Intelligence Layer , where analytics and insights can be templatized, visualized, and automated.

LiveRamp Clean Room's architecture lays the foundation for LiveRamp Clean Room's question framework, which drives insights consumed from clean room outputs.

mceclip0.png

Key Terminology

The following terminology will help you get up to speed on navigating the LiveRamp Clean Room UI:

  • Clean room: Secure, protected environment that allows multiple parties to bring data together for joint analysis in a privacy-compliant way.

  • Data connection: Used to access data at its source to minimize data movement.

  • Question: A business question in the form of an SQL query that can be reused for various runs or run-time parameters.

  • Question run: Question processing based on run-time parameters, such as dates and attributes.

  • Insights: Visualizations of question run results based on run-time parameters.

Roles and Responsibilities

As a partner collaborating in a clean room, there are several important steps required to properly onboard your data, often requiring members of key teams across your organization. Collaborations work best when people are aligned to roles across the clean room partner, clean room owner, and LiveRamp.

To get started, identify people in your organization that align best with the following clean room partner roles shown in the table below.

Roles

Responsibilities

Clean Room Partner

Account and Organization Admins

  • Clean room super-user administrator on behalf of partner

  • Creates users within LiveRamp Clean Room for colleagues; decides permission access tiers

  • Coordinates with IT to get the right data connections set up and configured

  • Advises the Marketing and Analytics teams on running reports within LiveRamp Clean Room

IT / Data Team

  • Supplies the credentials to the source system being integrated and provides context on the data itself, such as cloud, region, entity ID, and taxonomy.

Marketing / Marketing Analytics

  • Uses reports to enhance marketing and media performance

Account and Organization Admins

  • Supports onboarding process

  • Provides advisory support in analyzing query results

  • Consults on leveraging clean room analyses to answer key business questions

LiveRamp

Client Success Manager

  • Your LiveRamp business owner who will guide you through the LiveRamp Clean Room process and help you gain value out of the product

Technical Solutions Manager

  • Technical owner in charge of implementation, ongoing technical support and troubleshooting

Overall Implementation Steps

Getting started with LiveRamp Clean Room for publishers typically involves the following overall steps:

For more information on performing these overall steps, see the sections below.

Required Agreements

When adopting LiveRamp Clean Room, LiveRamp requires their Clean Room publishers to agree and sign the following contracting documents between the publisher and LiveRamp:

  • LiveRamp master services agreement (MSA) with clean room partner requirements

  • LiveRamp Data Processing Addendum (DPA)

  • LiveRamp statement of work (SOW) with addendum A, which offers a standardized legal framework for collaboration (Quick-Start Insights).

For advanced data collaboration use cases and features, publishers often require their own terms and conditions to be agreed upon with the brands. Those agreements need to be agreed between the parties before the collaboration.

Key Implementation Decisions

After you've signed the required agreements, your LiveRamp representative will work with you on the implementation decisions listed below, based on your situation and business needs.

Choose Clean Room Offerings

Establish your technical integration with LiveRamp Clean Room once to collaborate frictionlessly across the LiveRamp network and use any of our clean room-powered offerings. This includes:

  • Code-based workflows for data science and technical teams, which connects a brand and an individual partner (such as a publisher, CMN, or data seller) for custom collaboration (sometimes referred to as “1:1 custom collaboration”).

  • No-code access via LiveRamp’s Intelligence layer, designed for business users with pre-built queries and dashboards to accelerate time to insight. Quick Start offerings, which are part of the Intelligence layer, include:

    • Media Intelligence: Connects a brand and an individual publisher

    • Retail Intelligence: Connects a brand and a RMN (plus optional publisher overlay)

With this integration specification, we hope to simplify how you work with LiveRamp and set you up for scalable success - to meet your own business goals and make it easier to collaborate with your partners.

Choose Where Data Will be Hosted

One of the first decisions to make when implementing the LiveRamp Clean Room is identifying where the data will be hosted:

  • Connecting to data at the source (such as your cloud warehouse)

  • Connecting to data you’re already sending to LiveRamp

  • Sending the relevant data to LiveRamp to host

This decision is not mutually exclusive of where identity resolution happens. For some clients, identity resolution may happen in the LiveRamp-hosted environment. However, the data can still be connected at the source. Your account team can assist with tradeoffs between these scenarios.

Option

Details

Benefits

Direct connection to the data source

Data connections can be configured to any cloud-based storage location, including AWS, GCS, Azure Blob, Snowflake, Google BigQuery, and Databricks. For more information, see "Cloud-Based Data Connections".

It can provide flexibility, specifically with data updates, deletion of data connections, and in troubleshooting.

LiveRamp hosting

Sending data to LiveRamp can be executed through SFTP or setting up a connection to your S3 or GCS buckets. For more information about sending data to LiveRamp, see "Getting Your Data Into LiveRamp".

For customers with existing data feeds with LiveRamp, this offers a seamless way to get started with LiveRamp Clean Room.

If you choose to connect to your data at source, you’ll need to gather the following prerequisite information:

Prerequisite

Notes

Identify where your data is stored and in what region

LiveRamp Clean Room is interoperable and supports all major clouds, including:

  • Snowflake

  • Amazon AWS

  • Google Cloud Storage

  • Microsoft Azure

Identify the credentials associated with your cloud instance

For example:

  • Snowflake Account ID, Account Name, Organization Name & Region 

  • AWS S3 Location and File Type

  • GCS Project ID and Credential JSON

  • Azure SAS Token, Storage Account Name and Blob Directory

Identify the region, file format, and file path associated with your configuration

For example:

  • Schema & Table name

  • S3 / storage file path

  • Blob path

Identify the  dataset types you will be bringing into LiveRamp Clean Room

For example:

  • Identity data (i.e. ID graph)

  • User data (i.e. CRM, loyalty, audience segment)

  • Event data (i.e. transaction, ad logs)

  • Metadata (i.e. product data, campaign taxonomy)

For information on formatting each dataset type, see the “Data Formatting for Collaboration” section below.

Identify the user identifier being used for matching in LiveRamp Clean Room

Choose How to Incorporate Identity Resolution

Once you’ve determined how to connect your data to power LiveRamp Clean Room collaboration, the next step is configuring your identity resolution workflow so that your data can be mapped to RampIDs, LiveRamp’s universal, pseudonymous identifiers.

Connecting your data in LiveRamp Clean Room with RampIDs as the join key maximizes the value, ensuring the highest fidelity view can be connected across various identifier types across your partners. For more information on RampIDs, see “RampID” and “RampID Methodology”.

LiveRamp has two recommended options for identity resolution (depending on your goals) and will guide you through the process of deciding which will work best for you:

  • LiveRamp SFTP: Send the files you prepared to LiveRamp. Your account team will provide SFTP technical details. There, LiveRamp will apply identity resolution and connect the data to your LiveRamp Clean Room account.

  • Embedded Identity Resolution in your Environment:, Should you have an existing or preferred data environment, we offer cloud-based identity resolution in Snowflake, AWS, and BigQuery. Contact your account team for more details or see "Embedded Identity in Cloud Environments".

When preparing data for collaboration, consider the following recommended dataset structure for most data owners:

  • Universe data: This represents your full audience and likely includes all user identifiers (PII touchpoints or online identifiers) that will need to be used to resolve to RampIDs.

    Note

    When sending PII, it's important that as many PII touchpoints as possible are provided for LiveRamp's identity resolution capabilities to yield the best results.

  • Event / Impression data: Impressions or other data relevant for clean room collaboration with your partners.

All datasets should include a field for a CID (a unique customer identifier you leverage for deduping individuals across your datasets):

  • For PII-based universe datasets, we recommend that plaintext CIDs (not hashed) be sent (if you must hash the CIDs, we ask that you use MD5 hashing for interoperability).

  • For all other datasets (including universe datasets based on online identifiers), CIDs can be plaintext or hashed with any of our allowed hashing types (SHA-256, MD5, or SHA-1) but should be formatted consistently across datasets.

Note

If you use a combination of LiveRamp (SFTP) and Embedded Identity (Cloud Hosting), or if you send different dataset types, you must ensure that any hashed data is hashed using MD5 hashing in all files.

Identity Resolution Option

Datasets

Inputs

Identity Resolution Results

Update Cadence

LiveRamp (SFTP)

Universe data

  • Plaintext CIDs

  • Known (PII) identifiers (Name, Address, Email, Phone) or online (pseudonymous) identifiers

  • MD5-hashed CIDs

  • RampIDs

This resolved view will be applied to exposure data as it comes through

For most publishers, this data set will need to be refreshed monthly

Event / Impression data

  • MD5-hashed CIDs

  • Event data

  • MD5-hashed CIDs

  • RampIDs

  • Event data

If sending data through SFTP, send frequent updates for use in the LiveRamp Clean Room

Embedded identity (cloud hosted)

Universe data

  • Plaintext CIDs

  • Audience identifiers (Name, Email, Phone, etc.)

  • MD5-hashed CIDs

  • RampIDs

For most publishers, this data set will need to be refreshed monthly

Event / Impression data

  • MD5-hashed CIDs

  • Event data

N/A, the event data can be used in the data connection directly

If connecting to data at source, ensure the data connection will update with new impression information

To determine the best data hosting and identity resolution option for your business, you may need to mix and match solutions. For custom needs, including Local Encoder customization, consult your account team.

Choose How to Partition Large Datasets

LiveRamp enables you to indicate partition columns for your data connections in order to optimize query performance. Data partitioning is the process of dividing a large dataset into smaller, more manageable subsets. For partitioned data connections, data processing occurs only on relevant data during question runs which leads to faster processing times.

As a best practice, we recommend you partition the following fields for publisher measurement use cases:

  • Exposure date or impression date (for exposure or impression data)

  • Date or timestamp (for transaction or conversion data)

  • Brand (if applicable)

For more information, see "Data Connection Partitioning".

Set Up Your Account

Once you've made the necessary implementation decisions and your LiveRamp Clean Room organization has been created, a user with the Account Admin role can start setting up your account. This involves setting up permission sets via Roles and adding Users. Roles are the different combinations of clean room access permissions you can create and assign individual users to. When you create Users, you must assign them to a role.

Each newly created user will receive a welcome email to log into LiveRamp Clean Room. Their access level will mirror the role you have assigned them.

Note

Adding users to an organization does not mean these users have access to a configured clean room. Adding users to a clean room is a separate step.

Set Up SSO

To enable SAML-based SSO for LiveRamp Clean Room console access to your account, follow the instructions in "How to Setup SSO for LiveRamp Clean Room".

Manage User Roles

As part of LiveRamp Clean Room organization configuration, account and organization administrators can manage and add new user roles. Roles are combinations of Clean Room permission sets that can be assigned to individual users within an organization.

Clean Room provides a set of optional default roles that can be used as a starting point when setting up your permission structure. To access default roles, from the navigation pane select AdminRoles. Default roles can be edited or deleted.

Note

Suggested roles for consideration include Org Admin, Analytics Creator, Analytics View, Marketing Manager and Data Science. You can set up as many roles as you’d like.

For more information, see “Managing User Roles”.

Add Users

Once your account administrator has access to the LiveRamp Clean Room UI, they can create additional users. There is a separate flow for adding users to the UI versus adding users to a clean room.

For more information, see "Managing Users".

Test in a Sandbox Environment with Synthetic Data

To help users get real hands on keyboard skills within the Clean Room UI, LiveRamp will provide a sandbox environment to serve as a mock environment for experimentation. This clean room will consist of sample data and sample queries to help guide your team to learn the platform.

Your LiveRamp Client Success Manager will work with you on the testing process.

Format Data for Collaboration

LiveRamp offers a recommended data schema for clean room collaboration. The fields listed in the schemas below can be used to produce measurement results for the following clean room use cases (these will have pre-written queries for your use):

  • Audience Analytics

    • Overlap of brand data against publisher data

    • Audience Profiles

  • Media Delivery

    • Reach / Frequency

    • Impressions delivered across audiences, channels, platforms, campaigns, etc.

  • Attribution

    • Conversion counts

    • Lift

See the sections below for information on formatting your data, including the recommended schema and format for each dataset type you might need to connect to LiveRamp Clean Room. 

Note

While not all fields might be required for your current workflows and collaborations, we recommend you include as many of these additional fields now to make it easier for you to expand your collaborations in the future without you (or your IT team) having to perform additional work on your datasets.

The recommended field names shown aren’t mandatory, but using the recommended names can reduce the number of transformations you and your collaborators might need to do during dataset provisioning or query writing.

All data files or tables in a data connection job must have the same schema in order to successfully process. If you have multiple types of schemas, create a new data connection for each one. The order of the columns should match the schema.

Format a Universe Dataset

The universe dataset should represent your full audience and should include all user identifiers (PII touchpoints or online identifiers) that will be used during identity resolution to resolve to RampIDs. 

LiveRamp uses this dataset to create a mapping between your CIDs and their associated RampIDs. This mapping lives in a linked dataset and allows you to use RampIDs as the join key between the various datasets in queries.

For information on formatting and hashing identifiers, see “Formatting Identifiers”.

Note

  • When sending PII, it’s important that as many PII touchpoints as possible are provided for LiveRamp’s identity resolution capabilities to yield the best results.

  • Your CRM dataset might also be able to function as a universe dataset.

Field Contents

Recommended Field Name

Field Type

Values Required?

Description/Notes

A unique user ID

cid

string

Yes

  • LiveRamp uses the values in this field to resolve your data to RampIDs.

  • Plaintext CIDs are preferred. If you choose to hash the CIDs, make sure to use the same hashing type when sending CIDs in other data files or tables.

Consumer’s first name

first_name

string

Yes (if Name and Postal is used as an identifier)

Consumer’s last name

last_name

string

Yes (if Name and Postal is used as an identifier)

Consumer’s address

address_1

string

Yes (if Name and Postal is used as an identifier)

Consumer’s additional address information

address_2

string

No

  • Include values in this column if you have additional street address info for a given row.

Consumer’s city

city

string

Yes (if Name and Postal is used as an identifier)

Consumer’s state

state

string

Yes (if Name and Postal is used as an identifier)

  • Must be a two-character, capitalized abbreviation ("CA", not "California" or "Ca").

Consumer’s ZIP Code or postal code

zip

string

Yes (if Name and Postal is used as an identifier)

  • ZIP codes can be in 5-digit format or 9-digit format (ZIP+4).

Consumer’s best email address

email_1 

string

Yes (if email is used as an identifier)

  • Can be plaintext or one of our allowed hash types (SHA-256, MD5, or SHA-1).

  • If you have multiple emails for a consumer, send your best one in the “email_1” column.

Consumer’s additional email address

email_2

string

No

  • Can be plaintext or one of our allowed hash types (SHA-256, MD5, or SHA-1).

Consumer’s additional email address

email_3

string

No

  • Can be plaintext or one of our allowed hash types (SHA-256, MD5, or SHA-1).

Consumer’s additional email address

email_4

string

No

  • Can be plaintext or one of our allowed hash types (SHA-256, MD5, or SHA-1).

Consumer’s best phone number

phone_1

string

Yes (if phone is used as an identifier)

  • Do not include any hyphens or parentheses.

  • Can be plaintext or SHA-1 hashed.

  • If you have multiple phone numbers for a consumer, send your best one in the “phone_1” column.

Consumer’s additional phone number

phone_2

string

No

  • Do not include any hyphens or parentheses.

  • Can be plaintext or SHA-1 hashed.

Consumer's mobile device ID (MAID)

maid

string

Yes (if MAIDs are used as identifiers)

  • Can be plaintext or SHA-1 hashed.

Format an Exposures Dataset

An exposures dataset should Include data on impressions served or other ad events, along with metadata on the impressions or events.

Note

If you do not have values for any fields in this schema, we recommend that you still include those fields as placeholders.

Field Contents

Recommended Field Name

Field Type

Values Required?

Description/Notes

A unique user ID

cid

string

Yes

  • LiveRamp uses the values in this field to resolve your data to RampIDs.

  • If you sent plaintext CIDs in your universe dataset, send MD5-hashed CIDs.

  • If you hashed the CIDs in your universe dataset, use the same hashing type.

UTC timestamp of the event

event_timestamp

timestamp

Yes

  • (such as “2025-09-18T02:33:03Z”).

Unique ID for the creative served

creative_id

string

No

  • Values are required for this field for Meta TEE.

Name of the creative

creative_name

string

No

Unique ID for the placement, ad group, adSet or line item ID

placement_id

string

No

  • Values are required for this field for Meta TEE.

Name of the placement, ad group, or line item

placement_name

string

No

  • This is used as the most granular reporting metric, such as a group of impressions that is unique to a property or creative.

  • Values are required for this field for Meta TEE.

Unique ID for the campaign

campaign_id

string

No

  • Values are required for this field for Meta TEE.

Name of the campaign

campaign_name

string

No

  • Values are required for this field for Meta TEE.

Unique ID of the publisher or site where the exposure occurred

property_id

string

No

Location on a publisher site, the channel, or media type where the exposure occurred

property_name

string

No

Segment ID that the user belonged to at the time of exposure

segment_id

string

No

Unique ID for the advertiser account

account_id

string

No

  • Values are required for this field for Meta TEE.

Name of the advertiser account

account_name

string

No

  • Values are required for this field for Meta TEE.

Unique ID for the brand

advertiser_id

string

No

Name of the brand

advertiser_name

string

No

  • Values required for Quick-Start Insights, Cross-Media Intelligence, and Meta TEE.

Type of placement or line item

placement_type

string

No

Type of event

event_type

string

No

  • Possible values are “Impressions”, “Click”, or “View”.

  • Values required for Quick-Start Insights.

Cost of the event to the advertiser

cost

double

No

  • Values required for Quick-Start Insights.

Unique identifier for the event

event_id 

string

No

  • This field is very useful when CID to RampID mapping results in a 1 to many.

  • Values required for Quick-Start Insights and Cross-Media Intelligence

Format a Publisher Audience Dataset

For publishers, a publisher audience dataset includes data on user segments in your platform.

Field Contents

Recommended Field Name

Field Type

Values Required?

Description/Notes

A unique user ID

cid

string

Yes

  • LiveRamp uses the values in this field to resolve your data to RampIDs.

  • If you sent plaintext CIDs in your universe dataset, send MD5-hashed CIDs.

  • If you hashed the CIDs in your universe dataset, use the same hashing type.

Unique identifier of audience segment used for targeting this impression

segment_id

string

The name of the segment

segment_name

string

The name of the category

segment_category

string

Connect Your Data

Once you've confirmed that you have the required datasets ready for use, you're ready to connect your data. If you’re going to be hosting your data at its source in a cloud warehouse, you’ll need to connect that data to your Clean Room account.

Note

  • If you’ve chosen to have LiveRamp host the data, see the information in “Uploading Data”.

  • Before configuring your data connections in LiveRamp Clean Room, confirm you have the required datasets ready for use in a clean room. Knowing where your datasets live, the credentials and tables required to access those datasets, and having the proper file and table formatting in place will make the remainder of your setup much more seamless.

Connecting your data involves creating a data connection by performing the overall steps listed below. The type of connection to create will depend on your situation and business needs. Your LiveRamp representative will work with you to determine the type(s) of connections to create.

For more information on the steps to perform for your cloud provider and clean room type, see the articles in “Cloud-Based Data Connections”. Specific instructions for each step will be listed in the appropriate help article for your data connection type.

A data connection needs to be created for each dataset you want to utilize in a clean room.

After you create data connections, these connections appear on the Data Connections page, and you can create and configure clean rooms.

LCR-Getting_Started_Publishers-Data_Connections_page.png
Perform Actions in the Cloud Provider UI

For certain data connections, you’ll have to perform tasks in the cloud provider’s UI. This often involves things like downloading credentials, creating service keys, granting permissions, generating a token, or other similar actions.

Add Credentials

For LiveRamp to be able to access your data at your cloud provider, you’ll need to create a credential within LiveRamp Clean Room for that cloud account. This often involves entering a credential or token you generate from your cloud provider.

Create a Data Connection

Once you've created the appropriate credentials, you can create the data connection itself. During this process, you’ll also specify things like the file path LiveRamp should use to access your data, the file format, the field delimiter, and the data location.

Map the Fields

Once you’ve created the appropriate data connection and LiveRamp has successfully connected to your data, you then map and configure the fields. This includes specifying which fields you want to be queryable, any updates to column labels, which fields contain PII, and which fields should be used as identifiers.

Create and Configure Clean Rooms

After you've connected your data, you're ready to create and configure clean rooms by performing some or all of the overall steps listed below (depending on your situation).

For instructions on performing these steps, see the links above and the following articles:

Invite Clean Room Partners

Facilitating clean room collaborations requires clean room owners to invite partners to a clean room to combine and analyze data sources from multiple parties.

Note

Any partner you want to invite must have already set up their LiveRamp Clean Room account. Send the invitation to a user with Organization Administrator permissions who can then accept the invite on behalf of the partner organization in the clean room.

For instructions, see “Clean Room Partner Invitation Guide”.

Collaborate on Clean Room Questions

Once you and your partners are up and running in the clean room, you're ready to collaborate. See the information in the "Collaboration Guides" section of our documentation site for articles to help you with collaboration activities.