Guide
What is a canonical data model?
A canonical data model is a shared definition of data across an organisation's systems: names, attributes, characteristics, relationships and business meaning. "Canonical data model" and "common data model" usually mean the same thing. A data dictionary documents the terms and fields within it.
Why does a common data model matter?
Without a shared model, every integration, report and AI tool decides for itself what a customer, account or transaction is. The answers drift apart, and teams spend their time reconciling numbers instead of using them.
A common data model gives every downstream use the same starting definitions.
Example: two systems, one customer
Illustrative example. A wallet platform stores a customer as account_holder with fields acct_no, given_nm and status_cd. A marketplace stores the same person as member with member_id, full_name and is_active.
A proposed shared object might be person_records with attributes such as person_record_id, full_name, email_address and record_status, mapped from both sources.
- Is acct_no or member_id the identifier, or does the shared model need its own key?
- Does status_cd "S" (suspended) mean the same as is_active = false?
- Should two accounts held by one person become one record or two?
Who decides what the model says?
People do. Tools can propose structure and definitions from the evidence in source systems, but the questions above need business judgement. In Neuroplastiq they are flagged for review and the model only changes when someone approves the change.
Related guides
Content owner: Gareth Sutton. Last reviewed .