Please adhere to our file formatting guidelines when sending your data in the clean room to ensure efficient processing and highest possible match rates for your files.
How can you send the data?
Load your data into a table in your own Snowflake account, then grant the Resonate app SELECT access to it (see Step 2 of the Setup Guide). Resonate never has access to that data.
Best Practices for Optimal Match Rates
Reference File Format Example
Reference the file format example below, which shows the required columns and formatting requirements in the correct order. Following this format ensures your file aligns with the clean room's ingestion standards and avoids processing delays. If you do not have a certain ID type (e.g., MAID), leave that column out.
| CUSTOMER_ID | HEM | HEM_TYPE | MAID | IP_ADDRESS |
|---|---|---|---|---|
| CUST_001 | 051f26bd6cde782239bf52e56854d3feeca75ae5a84508d1ff9a1868ba167ee5 | HEMSHA256 | 3b4c1a2d-5e6f-7890-abcd-ef1234567890 | 203.0.113.45 |
| CUST_002 | NULL | NULL | 6f1e2d3c-4b5a-6978-8a9b-0c1d2e3f4a5b | 198.51.100.22 |
| CUST_003 | NULL | NULL | NULL | 192.0.2.10 |
Run the Hashing Script
Resonate developed a hashing script that standardizes email formatting prior to hashing:
- Converts all emails to lowercase
- Removes leading and trailing spaces
- Ensures clean, hash-ready input
Review Email Hashing and Data Cleanliness
To maximize match after running the hashing script, review your file to verify:
- All emails are lowercase
- No leading or trailing spaces
- All hash values are lowercase. Uppercase hashes pass validation but return no attributes.
After hashing, ensure hashes are valid alphanumeric outputs. Compare your hashes to the examples in the table below. For reference, three common hash types and their specifications:
| Hash Type | Length | Example |
|---|---|---|
| MD5 | 32 characters | 5d41402abc4b2a76b9719d911017c592 |
| SHA-1 | 40 characters | a94a8fe5ccb19ba61c4c0873d391e987982fbbd3 |
| SHA-256 | 64 characters | 051f26bd6cde782239bf52e56854d3feeca75ae5a84508d1ff9a1868ba167ee5 |
This table does not cover every ID type. See Column Formatting Rules by Identifier Type in Resonate Enrichment On-Demand Snowflake Clean Room for full specifications by ID type.
It is also worth reviewing your file for general data cleanliness:
- Remove duplicate, inactive, or low-quality records
- Suppress any rows missing critical IDs (e.g., customer ID + HEM, HEM_TYPE, or MAID)
- Remove any rows containing missing key fields or lacking consent (e.g., improperly hashed emails)
See Best Practices for Data Preparation for detailed advice on data cleanliness.
Required Identifiers for Matching
Resonate uses a waterfall matching method, starting with the most precise identifier and cascading to others when necessary. This strategy delivers the highest match possible, even when some IDs are unavailable.
You must include Customer ID and at least one of the following identifiers. Zip11 and name/address are not currently supported for clean room.
| ID type | Required fields | Notes |
|---|---|---|
| HEM + Customer ID | HEM, HEM_TYPE | Both columns required. HEM_TYPE must be exactly HEMMD5, HEMSHA1, or HEMSHA256. |
| MAID + Customer ID | MAID | Single field |
| IP Address + Customer ID | IP_ADDRESS | Supports IPv4 format |
Match Rate Scenarios
Resonate delivers market-competitive match rates, which vary based on the combination and type of identifiers you provide. See Fill Rates, Match Rates, and Waterfall Matching for guidance on reaching and optimizing match rates.
Need Help?
If you are unsure how to format your file or have any questions, contact your Customer Success Manager or email resonatesupport@resonate.com before submitting. We are happy to help ensure your data drives the best possible outcome.
Comments
0 comments
Article is closed for comments.