Sample Data Requirements

To best evaluate your data products, you must send a sample that most accurately showcases your data.

In order to get you onboarded smoothly, please follow all sample requirements.

General Guidelines

  • Your sample should include at least a 24 hour cycle of your overall data collection
  • Sample's should be between 100k - 500k total records or rows (by data type)

Format

Sample files must be submitted in .csv format

Please Note: zip files are not accepted at this time

Standard Schema

All required and optional columns must be in standard format and structure corresponding to the data package.

Please Note: your sample files must be in our standard schema to be properly evaluated.

Files without headers and/or required columns will need to be re-delivered from your team.

File Naming

All files must follow our naming guidelines to be properly ingested
  • File Name: <YYMMDD>_datatype.csv

SAMPLE DELIVERY LOCATION

AWS S3 Buckets

Please send samples to the bucket that was provided to you by your Partner Success team member, it should look like the following:

  • s3://narrative-from-[company]/sample

If your bucket was created before 2018, it will be named with “integrations” instead.

  • s3://narrative-integrations-[company]/sample

 

Please upload your data broken down by data type:

  • s3://narrative-from-[company]/sample/[datatype]/
    • s3://narrative-from-[companyname]/sample/age/
    • s3://narrative-from-[companyname]/sample/devicec/
    • s3://narrative-from-[companyname]/sample/digitalc/
    • s3://narrative-from-[companyname]/sample/idmap/
    • s3://narrative-from-[companyname]/sample/gender/
    • s3://narrative-from-[companyname]/sample/location/
    • s3://narrative-from-[companyname]/sample/tv/
    • s3://narrative-from-[companyname]/sample/purchase/

SAMPLE DATA CHECKLIST

Data Files

  1. .csv format (non zip)
  2. Files named properly (<YYMMDD>_datatype.csv)
  3. Follows standard schema
  4. Includes headers
  5. Includes all required columns/fields

Data Upload

  1. Sent to proper S3 Bucket 
  2. Follows bucket nomenclature

 

If your sample data does not follow these requirements, the onboarding process will take longer than necessary.