In short: Once data is collected, every item in Zeal moves through four steps: Moderation, Review, Validation and Enforcement. Each completed step adds a check mark, so you can see at a glance how far an item has progressed.
Understanding the Zeal workflow
Zeal is built to handle large volumes of data without losing accuracy. Automation does the heavy lifting where the rules are clear, and people step in where judgement is needed, so infringing content is identified and actioned quickly and consistently.
The workflow starts once Detection has brought data into Zeal, and has four steps:
Step | What happens | Who does it | Check marks |
1. Moderation | Each item is labelled as legitimate or a potential infringement | Auto-Moderation rules and moderators | One |
2. Review | Labels are quality checked and corrected if needed | Quality Analysts | Two |
3. Validation (optional) | The client confirms what should be enforced | Clients or external reviewers | Three |
4. Enforcement | Takedown notices are sent and tracked | Enforcers and Auto-Enforcement rules | – |
1. Moderation
Moderation sorts incoming data, so you know straight away what’s safe and what needs action. Every item is given a label, either automatically by a rule or manually by a moderator.
What can be moderated
Posts: marketplace listings, social media posts and webpages.
Images
Accounts: sellers and usernames.
Websites: domains.
Auto-Moderation
Auto-Moderation applies your predefined rules to label incoming content before anyone reviews it. Rules can use keywords, price, image features, logo detection and other parameters. Each item is marked as either:
Legitimate/Clean: safe, with no infringement.
Potential Infringement: labelled with the type of infringement.
Manual moderation
Some cases need human judgement. Moderators apply the same rule logic by hand, so nuanced cases are still labelled accurately. They:
Scan the details: titles, descriptions, images, account details and metadata.
Look for the same triggers as Auto-Moderation rules: keywords, domains, imagery and behaviour, using their own judgement.
Assign a label: the infringement type, such as counterfeit, unauthorised resale, phishing or impersonation.
Mark the item: as Legitimate (safe, no infringement, or authorised) or Potential Infringement.
Flag risk: use potentially actionable labels, such as Suspicious, for items that aren’t a clear infringement but show risk indicators.
When a label is applied
When a label is applied, by a rule or a moderator, the workflow status changes from To Moderate to Moderated by, showing whether the item was handled automatically or by a specific user. The item then shows one check mark.
Learn more: Auto-Moderation · Building Auto-Moderation Rules · Labels
2. Review (Quality Check)
Review makes sure moderation decisions are right before anything moves on. Quality Analysts check both manually moderated and auto-moderated items, so labels stay accurate, consistent and in line with brand and platform guidelines.
What can be reviewed
Posts: marketplace listings, social media posts and webpages.
Images
Accounts: sellers and usernames.
Websites: domains.
What Quality Analysts check
Post and image labels
Product category detection
Logo detection results
Any label that was assigned incorrectly is corrected or updated.
Review actions
Check: marks the item as quality checked, without sending it for enforcement.
Check & Enforce: marks the item as quality checked and flags it for enforcement.
Once quality checked, the item shows two check marks.
Learn more: Reviewing (QA) with Tasks
3. Validation
Validation gives you the final say on what gets taken down. It follows Review and is usually done by clients or external reviewers, who confirm whether an item should move forward to enforcement.
Note: Validation is optional. If it isn’t needed, items go straight from Review to Enforcement.
What can be validated
Posts: marketplace listings, social media posts and webpages.
Images
Accounts: sellers and usernames.
Websites: domains.
Why validate
Confirm enforcement eligibility: make sure items selected for takedown meet your brand’s enforcement criteria and policies.
Apply your own judgement: bring in commercial, legal or regional knowledge that rules or internal review may not capture.
Approve or reject enforcement: signal clearly whether an item should go to takedown or be excluded.
Stay in control: see and steer enforcement decisions, so you can trust the workflow.
Validation actions
Validate: confirms the label has been applied correctly.
Validate and Enforce: confirms the label is correct and the item is infringing, and marks it for enforcement.
On Hold: pauses the item while you wait for clarification or next steps.
No Action: the item has been reviewed and needs no further action.
Once validated, the item shows three check marks.
Learn more: Validating with Tasks
4. Enforcement
Enforcement is where infringing content gets removed. Enforcers, or Auto-Enforcement rules, use the applied labels to start takedown actions.
What can be enforced
Posts: marketplace listings, social media posts and webpages.
Accounts: sellers and usernames.
Websites: domains.
How enforcement works
Prepare enforcement notices in the Enforcement section (notice view).
Submit takedown notices to marketplaces, websites or social platforms.
Track the status of each request to confirm removals.
Manage pushbacks and re-enforce where needed.
Learn more: Understanding Enforcement · Enforcement (Notice View) · Auto-Enforcement · Takedown Status Filter
Quick reference
Not sure where to go next? Start here.
If you want to… | Start here |
See what happens at each step | Understanding the Zeal workflow (above) |
Label items automatically | |
Check that labels are correct | Review (Quality Check) (above) |
Approve items before takedown | Validation (optional) (above) |
Send and track takedowns | Enforcement (above) |
Manage the workflow day to day |