After running Identity Resolution, several outputs are written to your data warehouse. In this article, we'll explore each output and how to use them.

The following tables are created in the Output Dataset you specified in Settings:

The Final Identity Table is automatically added as a Dataset in GrowthLoop, making it available for Audiences, Journeys, and Exports.

Final Identity Table

The Final Identity Table is your unified customer view—one row per unique identity with the best available data merged from all sources.

Schema

The Final Identity Table includes:

📘

Note

The Final Identity Table contains only the columns you configured—there are no system-generated timestamp columns. Timestamps (_id_res_created_at, _id_res_updated_at) are available in the matched table if you need to track when records were processed.

Example

SELECT 
  primary_key,
  email,
  first_name,
  last_name,
  phone,
  postal_code
FROM your_dataset.final_identity_table
LIMIT 5;

Value Selection

When the same person exists in multiple sources with different values, the Final Identity Table uses data from the highest-precedence source that has a non-null value.

For example, if your sources have precedence:

1. CRM (precedence 1)
2. Email List (precedence 2)
3. Orders (precedence 3)

And a customer has:

• CRM: email = "[email protected]", phone = NULL
• Email List: email = "[email protected]", phone = "5551234567"

The Final Identity Table will have:

• email = "[email protected]" (from CRM, highest precedence)
• phone = "5551234567" (from Email List, since CRM is null)

📘

Note

Precedence is applied per-column, not per-record. Each column gets its value from the highest-precedence source that has a non-null value for that column.

Profile Report

The profile report stores metrics from each identity resolution run. Each run adds one row to this shared table.

Schema

Example Query

SELECT 
  report_id,
  capture_date,
  source_records_total_count,
  source_records_matched_count,
  golden_table_record_count,
  match_rule_results
FROM your_dataset._id_res_golden_profile_report
WHERE report_id = 'your_config_name'
ORDER BY capture_date DESC
LIMIT 1;

Matched Table

The matched table is an intermediate output showing which source records were grouped together and which rule matched them.

Schema

Example Query

Find all source records for a specific identity:

SELECT 
  _id_res_source_name,
  _id_res_primary_key,
  matched_by_rule,
  email,
  phone
FROM your_dataset._id_res_matched_yourconfig
WHERE match_id = 'abc123';

This shows that:

• The CRM and Orders records matched via Rule 1 (likely email-based)
• The Loyalty record matched via Rule 2 (likely phone-based)
• All three now share the same match_id

Querying Your Outputs

Count Unique Identities

SELECT COUNT(DISTINCT primary_key) as total_identities
FROM your_dataset.final_identity_table;

Find Multi-Source Identities

Identities that appear in multiple sources:

SELECT 
  match_id,
  COUNT(DISTINCT _id_res_source_name) as source_count,
  ARRAY_AGG(DISTINCT _id_res_source_name) as sources
FROM your_dataset._id_res_matched_yourconfig
GROUP BY match_id
HAVING COUNT(DISTINCT _id_res_source_name) > 1
ORDER BY source_count DESC
LIMIT 10;

Match Rate by Source

SELECT 
  _id_res_source_name,
  COUNT(*) as total_records,
  COUNT(matched_by_rule) as matched_records,
  ROUND(100.0 * COUNT(matched_by_rule) / COUNT(*), 2) as match_rate
FROM your_dataset._id_res_matched_yourconfig
GROUP BY _id_res_source_name;

Unmatched Records

Records that didn't match any rule:

SELECT *
FROM your_dataset._id_res_matched_yourconfig
WHERE matched_by_rule IS NULL;

Questions or issues? Reach out to us at [email protected].