Resonate Append Deliverables -Legacy Process for HEMs or MAIDs Only

Please note, this article supports existing Resonate Append clients using HEMs or MAIDs. We are transitioning all existing clients to which this applies to an improved delivery process. You can use this article as a reference if you have not yet transitioned. If you haven't transitioned yet, please reach out to your Customer Success Manager or contact resonatesupport@resonate.com to learn more and get started. 

Resonate will provide three output files as the deliverables for Resonate Append. We recommend using your data science or analytics resources who can help ingest these files internally in your environment.

The output file location is the same as the method used to send us input data

1. If the input data was sent using Snowflake; the output file is available under “Data Sharing” section of the account used to share the input table

   • The definition file will be shared with you by your Resonate customer contact

2. If the input data was sent using AWS; the output file is available in the same folder used to send input data

3. If the input data was sent using Box; the output file is available in Box.

4. If the input file was sent using SFTP; the output file will be available via SFTP; the output file will be available in the same SFTP location.

Resonate Output File Deliverables

1a. Standard Append Output File

The Standard Append Output File will be in a .csv file format and will include 3 columns.

• Column 1 contains the list of matched IDs from the input file. 
  • This column includes your IDs that were matched to Resonate IDs.
  • Note: you will notice less IDs in the file as compared to what was originally sent. This is because we ONLY send back the IDs that were matched to Resonate IDs.

• Column 2 contains the Predictions in a comma separated list of survey value keys.
  • These survey value keys are attributes that are true for the ID in column 1.
  • Only survey value keys corresponding to true attributes are included. The false are excluded.
  • The number of attributes included in the file depend on the clusters or attributes you choose.
• Column 3 contains a comma separated list of Redacted survey value keys.
  • These survey value keys are considered sensitive attributes and are redacted from the file and we will not deliver them for individuals residing in restricted states. 
• The number of attributes included in the file depend on the clusters or attributes you choose.
• Other attributes that are not included in the Predictions or Redacted columns can be interpreted as False for the ID in column 1.

1b. Pivoted Append Output File

The Pivoted Append Output File will be in a .csv file format and will include a column for each attribute ordered.

• Column 1 contains the list of matched IDs from the input file.

  • This column includes your IDs that were matched to Resonate IDs.

  • Note: you will notice less IDs in the file as compared to what was originally sent. This is because we ONLY send back the IDs that were matched to Resonate IDs.

• Column 2 onwards contains the survey value keys.

  • A 1 in the column indicates that attribute is true for the ID in column 1.

  • A 0 in the column indicates that attribute is false for the ID in column 1.

  • A NULL value (empty string) indicates that attribute is sensitive and is redacted from the file for individuals residing in restricted states.

• The number of attributes included in the file depend on the clusters or attributes you choose.
•  

2. Key Definitions File

• The file provides definition of the survey value keys. It provides the human readable name of the attribute tied to the key

• The customer must map the keys from column 2 in the standard output file to “survey_value_keys” column in the key definition file

• The customer must map the keys from column 2 onwards in the pivoted output file to “survey_value_keys” column in the key definition file

• This file lists all the clusters or attributes purchased by the customer and all possible attribute values tied to it.

• Example: If the output file contains a key = “422753”, that means the ID containing this key is in the age group 18-24.

• The file also includes whether the question the attribute was derived from was a single-select or multi-select question 

3. Summary Report

This document provides a summary of the contents in your Data Append Output file. It includes the following details:

• Columns A to E: Display the full taxonomy of each attribute value.
• SURVEYVALUEKEY: The survey_value_key corresponding to each attribute, as reflected in the output file.
• DISTINCT_COUNT: Represents the total number of IDs in the output file that are true for the given attribute value.
  • Example: If the DISTINCT_COUNT for "gender = female" is 188, this means we have predicted 188 female IDs in the customer output file.
• PERCENTAGE: The percentage of records that are true for the attribute value in each row.
  • Example: If the PERCENTAGE for "gender = female" is 94%, this means that 94% of the records in the customer output file are predicted to be female.

Use this file to verify that your Output and Key definition files are correctly aligned. After ingesting the file from us and formatting it on your end, we recommend that you perform a QA check by comparing your formatted data with the metrics provided in the summary report. For example, confirm if the counts for demographic categories like males and females in your formatted data align with the values in the summary report. Since sensitive data will be redacted for individuals in restricted states, percentage values in the summary report for mutually exclusive answers may no longer sum to 100% where previously they did. 

Joining the Output File Back to Your Customer Database

Resonate output files are easily ingested with the help of your analytics or data science team. Due to large file size and quantity of data, these files require more advanced technical resources and are difficult to interpret and analyze on tools like excel.

How to Read the Output Files

Below we’ve provided a simplified example to show how to interpret the file outputs.

In this example, your standard output file contains the following: