Data Co-op Update Logs - Issue #26

Data Co-op Update Logs - Issue #26
Photo by John Schnobrich / Unsplash

This is the first B2B/B2C delivery of 2023! The focus for these assets in 2023 is around quality and customization (more below).

Additional highlights for this update:

  1. Cookie Pool is officially in live beta!
  2. Added 17 million new B2B records and deprecated 38M (invalid) records
  3. Rollout has begun on custom quality parameters and delivery

Available Product Schemas

  • Cookie Pool 1.0 (beta)- Live: 2-1-23
  • B2C Contact Data 1.5 - Last Updated: 2-1-23
  • Full B2B Contact Data 4.5 - Last Updated: 2-1-23
  • IP>Company 2.6 - Last Updated: 2-1-23
  • Firmographic Data 1.7 - Last Updated: 2-1-23
  • MAID/HEM Extract 2.15 - Last Updated: 1-18-23
  • IP/MAID Extract 1.15 - Last Updated: 1-18-23
  • IP/MAID DA Extract 1.11 - Last Updated: 1-18-23
  • MAID/HEM DA Extract 1.11 - Last Updated: 1-18-23

Here are the highlights from the first full month of cookie tag fires:

  • Updated delivery frequency from once/day to <3 minutes
  • 868M 5x5 cookies placed
  • Realtime delivery (sub ~100 milisecond) is in testing

B2B 2.5.0 & B2C 1.5.0

Quality was the #1 focus for both B2B and B2C updates:

  • B2B: Added 17 million new B2B records and deprecated 38M invalidated records
  • B2B: A Graveyard file of deprecated (invalid) records is now available (please let us know if you're interested)
  • B2B: We have released Version 2.0 of the B2B quality models for improved hygiene

  • B2C: Improved email validation fill rate from 21M to 79M valid emails
  • B2C: Improved fill rate on demographic data such as martial status (7,900%), children (1,700%), net-worth (3,100%), and homeownership (4,044%)
  • B2C: The team is working on B2C 2.0 for March (more notes to come on that)

NEW: Custom Quality Models

What is a quality record? Ask 50 companies and you may get 49 different answers. As a result, we've built a new system that allows us to customize the delivery of data assets to ensure that they meet your specific quality requirements. Here is how the program works:

  1. Co-op members test the assets as they normally would and send the pass/fail data back to the Co-op
  2. The Co-op will then train its quality models to go find more data that looks like the pass data and less data that looks like the fail data
  3. From there, the Co-op sends back a custom data delivery uniquely created to solve your use case

Note: We're rolling this program out on a limited basis, please email us if you're interested in being a part of it

If you are subscribed to the B2B, B2C, Firmographic export or IP>Company, you'll receive a separate email with the bucket locations.

Please email if you have any questions regarding the most recent data delivery updates and if there are any data projects or issues that you’d like to see prioritized that we haven't mentioned.

We appreciate all of your feedback,

Nick Weldon, Founder of The Data Co-op