Cortex and Data: What You Need To Know

At the heart of Cortex is an Artificial Intelligence engine that powers all of the decision making for your automated campaigns. This AI is always hungry, and it's hungry for data.  If the engine doesn't get enough data or gets bad data, it fails to perform its jobs correctly.  Understanding what data Cortex needs, where the data come from, and how the data are processed is important for any marketer using Cortex.

What data does Cortex need?

Cortex needs two types of data: user data and item data.  User data includes demographics, on-site activity, and purchases.  Item data includes IDs, names, descriptions, and images.

How does Cortex get data?

Cortex gets data in two ways: on-site tracking and flat file transfers.  On-site tracking uses Javascript and API calls to send user activity back to your Cortex database.  Flat file transfers are standard batch files generated by your eCommerce platform, which are formatted like excel spreadsheets.  Add-to-cart and page view actions will come through JavaScript, while orders data will come through a flat-file.

What does Cortex do with my data?

Cortex uses your data to build models and uses those models to predict user preferences and run your automated email campaigns.  For example:

• Using customer information, Cortex is able to build a model of demographics and user actions that often lead to a purchase.  Users who fit into this model have "high purchase intent" and will receive stages like "New + High Intent" and "Ready To Buy."  
• Using item data and user data Cortex is able to build item preference models, to accurately target users with the products they are most likely to purchase 
• Using on-site tracking and order data, Cortex is able to see which users need to receive stages like Cart Abandon, Browse Abandon, and Follow-Up.
• Using item data Cortex knows when to send stages like New Arrivals and Item Back In Stock.