Required tables

The three core tables iCustomer Platform reads from your warehouse: Accounts, Contacts, and Events, plus the optional tables that enhance scoring and outcome tracking.

For warehouse workspaces.

The platform runs on three core data entities. Everything else is optional.

Table	Required	Purpose
Accounts	Yes	The companies you target
Contacts	Yes	The people at those companies
Events	Yes	The behavioral signals that drive scoring

Optional tables that make the system sharper:

Table	Required	Purpose
Opportunities	Optional	Pipeline data for outcome tracking
Products	Optional	Product-usage signals
Website Sessions	Optional	Anonymous visitor tracking

The minimum each table needs

• Accounts: account_id (unique), name, and domain. The domain is the matching key, so it matters most.

• Contacts: contact_id (unique), account_id (the link to the company), and email. Email drives identity matching.

• Events: event_id (unique), timestamp, entity_id plus entity_type (who did it), and event_type (what happened).

Fields beyond the minimum feed scoring directly. Industry, employee count, and revenue sharpen Fit. Title, department, and seniority map the buying committee. UTM fields on events carry attribution. Field types, examples, and warehouse DDL are covered when our team sets up your warehouse connection.

Blanks are not neutral. A missing domain or industry silently drops good accounts out of ranking, so completeness on the matching and Fit fields pays for itself.

Quality bar

Check	Target
account_id, contact_id, event_id unique	100%
account.domain present	above 95%
contact.email present	above 95%
event.timestamp present	100%
Accounts and Contacts refresh	daily (weekly minimum)
Events refresh	real-time or hourly (daily minimum)

Run the readiness checks in Preparing your data before connecting, and monitor after connecting in Data health.

Outside the app

Everything on this page can be inspected outside the app via API and MCP as well.