Why governance ensures data trust in dbt - Visual Breakdown

Concept Flow - Why governance ensures data trust

Define Data Governance Rules

↓

Apply Rules to Data Pipelines

↓

Monitor Data Quality & Compliance

↓

Identify & Fix Issues

↓

Ensure Data is Reliable & Trusted

Data governance sets rules, applies them to data, monitors quality, fixes issues, and ensures data trust.

Execution Sample

dbt

version: 2
models:
  - name: sales
    columns:
      - name: id
        tests:
          - unique
          - not_null
      - name: status
        tests:
          - accepted_values:
              values: ["completed", "pending", "cancelled"]

This dbt model config applies tests to ensure data quality and trust.

Execution Table

Step	Action	Rule Applied	Result	Data Trust Impact
1	Define tests in schema.yml	unique, not_null, accepted_values	Rules ready to apply	Foundation for trust
2	Run dbt tests on sales model	unique on id	Pass: all ids unique	Ensures no duplicates
3	Run dbt tests on sales model	not_null on id	Pass: no null ids	Ensures completeness
4	Run dbt tests on sales model	accepted_values on status	Fail: found 'in_progress'	Detects invalid data
5	Fix data issue: update 'in_progress' to 'pending'	Data correction	Pass on re-test	Restores trust
6	Monitor ongoing tests	Continuous checks	All tests passing	Maintains trust over time

💡 All tests pass after fixing invalid data, ensuring trusted data

Variable Tracker

Variable	Start	After Step 2	After Step 4	After Step 5	Final
Data Quality Status	Undefined	Unique test passed	Accepted values test failed	Data corrected	All tests passed

Key Moments - 3 Insights

Why is it important to define tests before running them?

Defining tests sets clear rules to check data quality, as shown in Step 1 of the execution_table.

What happens when a test fails during data validation?

A failure, like in Step 4, signals bad data that must be fixed to restore trust.

How does fixing data issues affect data trust?

Fixing issues (Step 5) makes tests pass again, which rebuilds and maintains trust.

Visual Quiz - 3 Questions

Look at the execution_table, what test failed at Step 4?

A unique on id

B not_null on id

C accepted_values on status

D no test failed

💡 Hint

Check the 'Rule Applied' and 'Result' columns at Step 4 in execution_table

According to variable_tracker, what is the data quality status after Step 5?

A Unique test passed

B Data corrected

C Accepted values test failed

D All tests passed

💡 Hint

Look at the 'After Step 5' column for 'Data Quality Status' in variable_tracker

If the data issue was not fixed at Step 5, what would happen to data trust?

A Data trust would be lost due to failing tests

B Data trust would remain high

C Data trust would improve automatically

D Data trust is unrelated to tests

💡 Hint

Refer to Step 4 failure and Step 5 fix impact in execution_table

Concept Snapshot

Data governance means setting rules to check data quality.
Use tests like unique, not_null, and accepted_values.
Run tests regularly to find bad data.
Fix issues quickly to keep data trusted.
Trusted data helps teams make good decisions.

Full Transcript

Data governance ensures data trust by defining clear rules to check data quality. In dbt, these rules are tests like unique, not_null, and accepted_values. The process starts by defining tests, then running them on data models. If tests pass, data is trusted. If tests fail, it shows data problems that must be fixed. Fixing issues makes tests pass again, restoring trust. Continuous monitoring keeps data reliable over time. This step-by-step approach builds confidence that data is accurate and ready for use.