How to Create Data Governance Rules for Naming Conventions

Automatically generate data governance rules that monitor your naming conventions for compliance, catching violations like missing parts, incorrect delimiters, or values outside your allowed lists.

Overview

When you define a naming convention in Improvado, you establish a structure for how campaigns, ad sets, or other entities should be named. But as your team scales, naming mistakes inevitably slip through.

Data governance rules bridge that gap. With a single click, Improvado generates a compliance rule for each dimension in your naming convention. These rules continuously scan your live data and flag any entity names that don't match the structure you defined — so you can catch issues before they affect reporting.

Rules are created and managed through the AI Agent chat interface. Once created, they appear on your Data Governance Dashboard alongside any other rules you've configured.

How It Works

Each data governance rule checks a single dimension (e.g., Campaign Name, Ad Set Name) against the structure defined in your naming convention. Specifically, the rule:

1. 1Splits each entity name by the delimiter you configured (typically <code>-).
2. 2Checks each required part against its allowed values list. If a part is empty or contains a value not in the allowed list, it's flagged as a violation.
3. 3Skips optional parts — only required parts are validated.
4. 4Applies filters — if your naming convention has filters configured (e.g., only check a specific advertiser), the rule respects those filters.

A single entity name can have multiple violated parts. The rule output shows exactly which parts failed and what values were found, making it easy to identify and fix the issue.

Create Rules

1. 1Open your naming convention from the Naming Convention page.
2. 2Click the Create Data Governance Rules button in the toolbar.
3. 3If rules already exist for this naming convention, a confirmation dialog appears asking whether you want to overwrite them. Click Confirm to proceed.
4. 4A new AI Agent chat opens and automatically generates one rule per dimension. You'll see the progress in the chat as each rule is created.
5. 5Once complete, your rules are live. Once the rules are processed, navigate to the Data Governance Dashboard to view them.

Each rule is named using the format: {Naming Convention Name} - {Dimension Column} (e.g., "Meta Ads Campaigns - campaign_name").

Default Rule Settings

When rules are first created, they use the following defaults:

You can modify any of these settings from the Data Governance Rules page after creation.

Automatic Rule Sync

When you update your naming convention and save changes, Improvado automatically checks whether the update affects your data governance rules. If it does, the rules are regenerated to match the new structure — no manual action required.

Changes that trigger an automatic sync include:

• Adding, removing, or reordering naming parts
• Changing a part's optionality (required vs. optional)
• Modifying the delimiter
• Updating filter columns

Changes that do not trigger a sync (because they don't affect rule logic):

• Editing allowed values within a part
• Updating workbook metadata that doesn't affect structure

When a rule is synced, any customizations you've made to its settings — such as owners, status, severity, or description — are preserved. Only the underlying compliance query is regenerated.

Rule Deletion

Automatic deletion

Data governance rules are automatically deleted when the associated naming convention or dimension is removed:

• Deleting an entire naming convention removes all linked rules.
• Deleting a single dimension removes only the rule associated with that dimension.

This cleanup happens automatically before the naming convention data is removed, ensuring no orphaned rules remain.

Manual deletion

You can also delete individual rules from the Data Governance Rules page at any time, just like any other data governance rule.

What Gets Checked

The rule queries your live source data directly, so violations are detected as soon as new data arrives.

For each entity name in the dimension, the rule evaluates every required part:

• Empty part — the delimiter split produced a blank value (e.g., a double delimiter -- or a trailing delimiter).
• Value not in allowed list — the part contains a value that isn't in your configured allowed values. If no allowed values are defined for a part, that part is not checked.