Activation Overlap Analysis

The Activation Overlap Analysis template creates an audience based on the matched overlap (intersection) of an Instacart table and a partner-provided table.

The template matches customer identifiers that are provided by partners, such as Hashed Email (HEM) and Hashed IP Address, with Instacart's user database to generate match statistics.

You can perform the following actions with this template:

• Match hashed customer identifiers to Instacart's user database.
• Work with multiple identifier types, such as hashed emails, hashed IP addresses, and hashed phone numbers.
• Share the overlapping audience with Instacart's Snowflake account allowing Instacart to push it to external advertising platforms on behalf of the partner, enabling activation within advertising campaigns.

Set up the activation policy

Before running the Activation Overlap Analysis template, set an activation policy to specify which columns from your consumer table can be sent to Instacart's account. You must approve all columns used in the Activation Overlap Analysis template, including join columns. This is a one-time setup for each clean room.

Run the following query to set an activation policy:

set cleanroom_name = 'YOUR_CLEANROOM_NAME';
call samooha_by_snowflake_local_db.consumer.set_activation_policy(
    $cleanroom_name,
    ['activation_overlap_analysis:YOUR_DB.YOUR_SCHEMA.YOUR_CUSTOMER_DATA:IDENTIFIER_TYPE',
     'activation_overlap_analysis:YOUR_DB.YOUR_SCHEMA.YOUR_CUSTOMER_DATA:IDENTIFIER_VALUE']
);

Each element in the array uses the following pattern: template_name: fully_qualified_table:column_name.

For more information, see Snowflake's Consumer Activation Policy Documentation.

Run the activation query

To run the activation query, use the run_activation procedure instead of the run_analysis procedure.

set cleanroom_name = 'YOUR_CLEANROOM_NAME';
call samooha_by_snowflake_local_db.consumer.run_activation(
    $cleanroom_name,                          -- Cleanroom name
    'activation_overlap_analysis',            -- Activation template name
    ['YOUR_DB.YOUR_SCHEMA.YOUR_CUSTOMER_DATA'], -- Your customer identifier table
    'segment_name',                           -- Segment name for this activation
    object_construct()                        -- Additional parameters (not required at this time)
);

Parameters

The following parameters are used in the activation query:

Parameter	Description	Data type	Required or optional
clean room name	The name of your clean room environment provided by Instacart.	string	required
template name	The name of the Activation template. Currently the only template is the following: - activation_overlap_analysis. Activates matched customer identifiers based on overlap analysis.	string	required
consumer table	A table in your Snowflake account containing customer identifiers to match. The table must include the following: - IDENTIFIER_TYPE. The type of identifier (HEM, HASHED_IP_ADDRESS or HASHED_PHONE_NUMBER). - IDENTIFIER_VALUE. The hashed identifier value. For more information, see Consumer table schema.	string	required
segment name	A string used to identify this activation run. This helps organize and track multiple activation campaigns. If you use the same segment name in multiple activation runs, results for those activations will get replaced with the latest run. For information about the naming convention, see Segment naming.	string	required

note

Template-specific parameters are passed as an object. Additional parameters are not required at this time.

Consumer table schema

Your consumer identifier table must follow this structure:

CREATE TABLE YOUR_CUSTOMER_DATA (
IDENTIFIER_TYPE VARCHAR, -- 'HEM', 'HASHED_IP_ADDRESS' or 'HASHED_PHONE_NUMBER'
IDENTIFIER_VALUE VARCHAR -- The hashed identifier value
);

For example:

IDENTIFIER_TYPE	IDENTIFIER_VALUE
HEM	a1b2c3d4e5f6789...
HEM	f6e5d4c3b2a1098...
HASHED_IP_ADDRESS	1a2b3c4d5e6f789...

Data requirements

When sharing or processing data, follow these data requirements:

• Hashed values only. Ensure all identifiers are pre-hashed using the agreed-upon hashing algorithm.
• Consistent hashing. Use the same hashing method across all identifiers. Coordinate with Instacart.
• Valid types. Ensure IDENTIFIER_TYPE is one of the following supported types:
  • 'HEM'
  • 'HASHED_IP_ADDRESS'
  • 'HASHED_PHONE_NUMBER'
• No PII. Never include raw email addresses, IP addresses, or other personally identifiable information.

Activation results

Instacart uses the activated identifiers to do the following:

• Create custom audiences using external advertising platforms.
• Enable partners to activate audiences for targeted advertising campaigns through an external advertising platform.
• Measure campaign effectiveness through subsequent analysis queries.

note

Instacart will work with you separately on campaign setup and activation strategy for external platforms.

Activation results are sent to Instacart's Snowflake account in a designated activation summary table. Partners do not have access to this table or the matched identifiers.

Instacart receives the following information for each activation:

• Matched identifier types and hashed values
• Segment name you provided
• Timestamp of the activation
• Clean room name and activation ID

For details about how Snowflake manages activation results, see Viewing Activation Results.

Use cases

You can use this template in the following ways:

• Targeted advertising campaigns. Activate your customer lists on external platforms through Instacart to run targeted advertising campaigns based on purchase behavior and preferences.
• Lookalike audience building. Use matched identifiers as seed audiences to build lookalike audiences on advertising platforms for customer acquisition.
• Re-engagement campaigns. Activate lapsed customers identified through clean room analysis for re-engagement campaigns on advertising platforms.

Best practices

Consider the following when using this template:

• Consistent hashing. Coordinate with Instacart on the hashing algorithm to ensure match accuracy.
• Quality data. Clean and normalize identifiers, such as make lowercase and remove spaces, before hashing.
• Current data. Use recent customer data for better match rates and campaign relevance.
• No test data. Remove internal test accounts and invalid identifiers before upload.
• Compliance policies. For more information, see Data privacy and security.
• Query execution timing. For more information, see When to run the query.
• Instacart collaboration. For more information, see Work with your Instacart representative.
• Segment naming parameters. The segment name (IC_DCR_YYYYMMDD_META_XX_XXXXXXXX_XXXXXXX) is a string you provide to identify this activation run. You can use the same segment name across multiple runs to replace existing results, or use different names to keep activations separate. For information about the naming convention, see Segment naming.
• Activation query completion. Activation queries may take several minutes to complete. Results are transferred and decrypted in Instacart's account after the query completes.
• No visibility to direct results. Unlike analysis queries, you don't see the matched identifiers or activation results in your account.