Compliance Rules
Compliance rules define the guardrails that Govern enforces across your content. Each rule targets a specific check type and can be configured with custom parameters.
Creating Rules
Section titled “Creating Rules”- Navigate to Govern from the sidebar
- Go to the Rules section
- Click Create Rule
- Configure:
- Name — a descriptive name (e.g., “Brand voice consistency”)
- Check type — brand consistency, legal compliance, terminology, or tone
- Severity — hard (must comply) or soft (should comply)
- Configuration — type-specific parameters (see below)
- Active — toggle to enable or disable the rule
Rule Types and Configuration
Section titled “Rule Types and Configuration”Brand Consistency
Section titled “Brand Consistency”Evaluates content against your Brand DNA Vault voice settings using semantic similarity scoring.
Configuration:
- Guideline ID — which brand guideline set to compare against
- Minimum similarity score — threshold for passing (e.g., 0.85 means content must be at least 85% similar to your brand voice)
Content that falls below the similarity threshold is flagged with suggestions for making it more on-brand.
Legal Compliance
Section titled “Legal Compliance”Checks for required disclosures and regulatory language.
Configuration:
- Required disclosures — list of disclosures that must appear in content (e.g., “Results may vary”, “Not financial advice”)
Content missing any required disclosure is flagged with the specific disclosure that needs to be added.
Terminology
Section titled “Terminology”Enforces your approved and prohibited term lists from Brand DNA Vault.
Configuration:
- Approved terms — words and phrases that should be used
- Prohibited terms — words and phrases that must not appear
Content using prohibited terms is flagged with the specific term and a suggestion for the approved alternative. Content missing preferred terms receives a soft warning.
Verifies content matches your expected formality and voice.
Configuration:
- Expected tone — formal, neutral, or casual
Content that doesn’t match the expected tone is flagged with an explanation of how it deviates and suggestions for adjustment.
Managing Rules
Section titled “Managing Rules”Enabling and Disabling
Section titled “Enabling and Disabling”Each rule has an active/inactive toggle. Disabled rules are not evaluated during compliance checks. This lets you keep rules defined for future use without enforcing them.
Severity Levels
Section titled “Severity Levels”- Hard — content cannot be approved until hard rule violations are resolved
- Soft — content can still be approved, but warnings are surfaced for consideration
Editing Rules
Section titled “Editing Rules”Update any rule’s name, configuration, or severity at any time. Changes take effect on the next compliance check run — previously evaluated content is not retroactively re-checked unless you trigger a new check.
- Start with terminology rules — they’re the most straightforward and catch common brand consistency issues immediately
- Set realistic similarity thresholds — 0.85 is a good starting point for brand consistency; lower thresholds produce fewer false positives
- Use hard severity sparingly — reserve it for genuine legal requirements; overuse creates friction in content workflows
- Review rule effectiveness — periodically check compliance results to see if rules are producing useful findings or just noise