Skip to main content
This is likely a permissions issue. You need access to workflows to create and activate them.Solution: Contact your workspace administrator to grant you workflow access under Settings → Roles.If you don’t see the Workflows section at all in your sidebar, this confirms it’s a permissions issue.
Manual workflows only appear in the navbar if properly configured:
  1. The workflow must be activated (not in draft mode)
  2. The navbar placement must be set to Pinned
  3. For Single/Bulk triggers, you must be on the correct object page
To check: Open the workflow → click the trigger → verify “Navbar placement” is set to “Pinned”.You can always access manual workflows via Cmd + K (or Ctrl + K) regardless of navbar settings.
TypeRecords RequiredWorkflow Runs
GlobalNoneOnce, no record input
SingleOne or more selectedOnce per selected record
BulkOne or more selectedOnce, with all records as array
  • Global: Use when the workflow doesn’t need any record context (e.g., generate a report)
  • Single: Use when you want to process each selected record independently (e.g., send individual emails)
  • Bulk: Use when you need to process records together or optimize credit usage (requires Iterator action)
See Workflow Triggers for details.
An explicit If/Else node is not yet available but is on our roadmap.Current workaround: Create multiple branches from your step, each starting with a Filter action:
Step 1
   │
   ├── Branch A: Filter (condition = true) → Actions...
   │
   └── Branch B: Filter (condition = false) → Actions...
Only the branch where the filter condition passes will execute its subsequent actions.See How to Use Branches for a step-by-step guide.
Yes, branches run in parallel by default.If you want only one branch to execute:
  • Add a Filter action at the start of each branch
  • Set opposite conditions (e.g., Branch A: status = “Open”, Branch B: status ≠ “Open”)
Branches that fail their filter condition stop executing, while others continue.
Yes. After your parallel branches complete, you can add a step that both branches connect to.In the workflow editor:
  1. Complete your branched actions
  2. Add a new step after the branches
  3. Drag connections from the end of each branch to this new step
The merged step will execute after all connected branches complete.
Search Records returns a maximum of 200 records.If you need to process more:
  • Add more specific filters to reduce results
  • Use scheduled workflows to process in batches
  • Consider using the API for bulk operations
For most workflows, 200 records is sufficient. If you regularly hit this limit, consider restructuring your automation.
Not yet. CC and BCC fields for the Send Email action are on our roadmap.Current workaround: Add multiple Send Email actions to send to additional recipients, or use an HTTP Request to send via an external email service that supports CC.
Every action produces output data that can be used in subsequent steps.To reference previous step data:
  • Use the variable picker when configuring a field
  • Or type {{stepName.fieldName}} directly
Examples:
  • Trigger data: {{trigger.object.email}}
  • Search results: {{searchRecords[0].name}}
  • Code output: {{code.calculatedValue}}
Hover over any field in the action configuration to see available variables from previous steps.
Iterator requires an array input. Common issues:
  1. Input is not an array: Ensure you’re passing results from Search Records or another action that returns an array
  2. Array is empty: Add a filter before Iterator to check {{searchRecords.length}} > 0
  3. Wrong variable selected: Make sure you select the array itself, not a single record
Correct setup:
  1. Search Records (returns array)
  2. Filter: length > 0
  3. Iterator: select {{searchRecords}}
  4. Actions inside iterator use {{iterator.currentItem.fieldName}}
Code actions (serverless functions) have a default timeout of 5 minutes (300 seconds).The maximum configurable timeout is 15 minutes (900 seconds).If your code exceeds this limit, the action will fail with a timeout error.Tips to avoid timeouts:
  • Break large operations into smaller chunks using Iterator
  • Avoid heavy computations; use external services via HTTP Request for intensive processing
  • Optimize your code to reduce execution time
  • If you need longer processing, consider using scheduled workflows that process data in batches
Workflow runs show the execution history and help you debug issues.Access runs:
  • In workflow editor → Runs panel on the right
  • Or go to Workflow Runs in the sidebar
Understanding a run:
  • Status: Running, Completed, Failed, Waiting
  • Steps: See which steps executed and their output
  • Errors: Click failed steps to see error messages
  • Data: View input/output data at each step
See Workflow Runs for details.
Workflow runs might be failing immediately due to rate limits.Hard limit: 5,000 runs per hour per workspace.If you exceed this limit, workflows are immediately marked as failed and won’t appear in your runs list as expected.Common scenarios that hit this limit:
  • Selecting more than 5,000 records with a Single manual trigger
  • Multiple workflows running simultaneously across your workspace
  • High-frequency automated triggers (e.g., Record Updated on a busy object)
Solutions:
  • Use Bulk triggers instead of Single to process many records in one run
  • Space out large batch operations
  • Use filters to reduce trigger frequency
  • Schedule heavy workflows during off-peak hours
Twenty has two rate limits to ensure system stability:
LimitValueBehavior
Soft limit100 runs/minuteRuns queue in “Not Started” status, processed gradually
Hard limit5,000 runs/hourRuns immediately fail
Soft limit (100/min): Your workflows won’t fail—they just wait in the queue and are processed over time. You can trigger more than 100 records; execution will be slower.Hard limit (5,000/hr): This applies to your entire workspace. If all your workflows combined exceed 5,000 runs in an hour, additional runs will fail immediately.Tips to stay within limits:
  • Use Bulk triggers with Iterator instead of Single triggers for large batches
  • Combine related automations into fewer workflows
  • Use scheduled workflows to spread load over time
No, there is no automatic retry functionality at the moment.If a workflow run fails, you’ll need to:
  1. Review the error in Settings → Workflows → [Your Workflow] → Runs
  2. Fix the issue (data, configuration, or external service)
  3. Manually trigger the workflow again on the affected record(s)
Tips to reduce failures:
  • Add Filter nodes to validate data before actions
  • Use Search Records to check if related records exist
  • Test thoroughly with a few records before bulk operations
Automatic retry functionality is on our roadmap for a future release.
Yes, if your workflows are triggered by record creation or updates.When you import data via CSV, each record created or updated can trigger workflows. A large import (thousands of records) could:
  • Hit the 5,000 runs/hour limit
  • Consume significant workflow credits
  • Send unexpected emails or notifications
  • Create duplicate tasks or records
Before a mass import:
  1. Go to Settings → Workflows
  2. Identify workflows triggered by the object you’re importing
  3. Deactivate them temporarily
  4. Run your CSV import
  5. Reactivate the workflows when done
Alternative: If you need the workflows to run on imported data, import in smaller batches to stay within rate limits.
If your workflow canvas looks messy with nodes scattered around, you can automatically organize it:
  1. Right-click anywhere on the workflow canvas
  2. Click Tidy up workflow
This will automatically rearrange all nodes into a clean, organized layout.