HubSpot workflow errors can mean an action failed, was skipped, is waiting for a retry, or was rescheduled for later. Those outcomes require different responses. Start in the action log, locate one affected record and one specific step, then classify the event before you edit the workflow or manually enroll the record again.
Quick answer: open Automation → Workflows → View details → Action logs. Filter to the relevant date, record, action, and workflow revision. Read the exact event message. A message that says “will retry soon” is different from a permanent data error; a rescheduled action may still run in its next allowed time window. For errors, use HubSpot’s Review automation issues view to separate problems with the workflow setup from problems with the enrolled record (HubSpot: Troubleshoot common workflow errors).
First classify what the log says
Do not treat every red or delayed event as the same failure. The action log tells you whether the record needs a data correction, a workflow edit, a wait, or a check in another system.
| Log outcome | What it usually means | First response |
|---|---|---|
| Error | The action could not complete under the current workflow or record state. | Open the full error detail and determine whether the cause is the workflow, the record, or a connected system. |
| Information / retry | HubSpot expects to try the action again, often after rate limiting, a timeout, or temporary service load. | Check the destination system and retry message before manually enrolling the record. |
| Rescheduled | A timing rule moved the action to a later eligible time. | Confirm the workflow’s allowed action window and the displayed next time. |
| Skipped | The action was intentionally bypassed or could not run in the current state—for example, the workflow was off or an email step was ineligible. | Find the skip reason and decide whether the record should receive that step at all. |
| Completed, but the outcome is missing | The workflow step may have completed while the expected email or downstream effect was not observed. | Check the branch taken, record eligibility, recipient mailbox, and downstream system logs. |
A workflow’s individual action history and the site-wide automation issue list serve different purposes. The first traces what happened to a record at a step; the second helps you find recurring issues across workflows and records.
Find the exact action and workflow version
- In Automation → Workflows, open the workflow’s details and select Action logs. Search for the contact, company, deal, or ticket that was affected.
- Set the date range around the incident. HubSpot displays action timestamps in the account’s default time zone, so compare the log with CRM, form, and integration timestamps using that same zone.
- Filter by event type and action. If several edits occurred, select the workflow revision for the time of the event; the current editor may no longer match the version that processed the record.
- Open the event details and copy the exact message, record ID, action name, and timestamp. For an issue affecting many records, compare a successful record with a failing one at the same step.
For a broader incident list, open the Workflows dashboard and choose Review automation issues, then filter to automation issues created in Workflows. HubSpot distinguishes workflow-configuration problems from record-specific problems in this view.
Common errors and the evidence that resolves them
An associated record is missing
An action may try to update a company, deal, or other associated record that the enrolled contact does not have. Check the record’s associations first. For an individual record, correct the association if it should exist. For future enrollments, add a guard so the action runs only when the required association is present. A filter that prevents an invalid action is safer than repeatedly retrying a record that still lacks its target.
A duplicate record blocks a create action
HubSpot prevents duplicate contacts and companies when a new record uses an email address or company domain that already exists. The create action can fail, and later actions that depend on the new record may fail as well. Check whether the matching object already exists. Then branch the workflow: associate or update the existing object when appropriate, and create a new one only when no match exists.
A lifecycle-stage update tries to move backward
HubSpot’s default lifecycle-stage property only advances down the funnel. An action that tries to change a record from a later stage to an earlier stage can fail. Confirm that the workflow’s business rule is correct before clearing and setting the property again; clearing a lifecycle stage can affect reports and handoffs that rely on it.
A workflow email was skipped or not received
Check whether the contact is a marketing contact, has a usable email address, and is eligible under the subscription and privacy settings for that email. A non-marketing contact can have the email step skipped while other workflow actions continue. Unsubscribes, hard bounces, and suppression for low engagement have separate causes. Do not override an opt-out; follow the contact’s subscription state and applicable consent rules.
A “sent” or “completed” event does not prove that the message reached the inbox. Verify which branch the contact followed, then check spam, mailbox rules, and the recipient’s address. Also review the email’s own delivery or suppression details.
The connected app or webhook timed out
A “retry soon” event can point to rate limiting, a slow response, or an error in the connected service. Check both systems using the same record ID and timestamp. A webhook may have performed its side effect but failed to return a response before HubSpot’s time limit; HubSpot may then retry it. Practical implication: before manually enrolling the record again, confirm whether the destination already processed the request. Where you control the receiving system, make repeated requests safe to handle by using a stable event or record key.
The action is waiting or rescheduled
HubSpot may queue actions when many records arrive at once or the account is handling heavy CRM activity. A workflow can also reschedule a step that arrives outside its configured action window. Read the exact next-action time before changing the rule. If a bulk enrollment is creating load, split it into smaller batches and stagger them instead of repeatedly restarting the same records.
An AI action is paused for credits
Some HubSpot AI workflow actions can stop when the account has no available credits or an administrator has paused credit usage. The log may indicate that HubSpot will retry on a schedule, but retries have a limit. Check account credit availability and limits before changing the workflow’s downstream property actions; an empty AI output can otherwise be written into a record.
A practical triage map
| Evidence | Likely cause | Next diagnostic step |
|---|---|---|
| “No associated target records” | Missing contact-company, deal, or other association | Inspect associations and add an eligibility condition before the action. |
| “Existing record shares a unique value” | Create action is trying to recreate a contact or company | Find the match and branch to associate/update versus create. |
| Email step skipped | Contact type, address, subscription, bounce, or suppression | Inspect that contact’s eligibility and the automated email’s settings. |
| “Retry soon” / rate limit / timeout | Temporary limit or slow connected service | Compare HubSpot’s log with the receiving app before retrying manually. |
| “Action rescheduled” | Allowed action-time window or queued processing | Check the next scheduled time and workflow timing settings. |
| No success event in the log | Date filter, revision, retention, or daily success-log cap | Adjust the date and revision filters; compare record history and the destination system. |
A missing log entry is not always proof of a missed action
HubSpot retains workflow action logs for 90 days and enrollment history for six months. It also documents a daily limit of 100,000 successful workflow-execution logs. If that success-log limit is exceeded, success and information events stop being stored for the rest of the day, while error logs continue to appear; records continue through the workflow. Therefore, an empty success log can reflect a logging limit or the selected date/revision filters rather than a stopped workflow. Check the record path and the downstream system before concluding that an action never ran.
Repair the cause, then recover only the affected records
- Fix the source condition: record value, required association, email eligibility, workflow timing, or connected app response.
- Check whether HubSpot will retry automatically. Avoid adding a second manual execution while the original action may still be queued or retrying.
- If a retry is needed, select only records whose action failed and whose side effect did not already occur. Review later workflow steps because a failed create or update can affect actions that depend on its result.
- Test the intended path with a safe record or the workflow preview, then inspect one real event and its downstream result before processing a larger group.
- Document the event message, cause, correction, and affected records. Mark an automation issue as fixed only after confirming the same failure no longer recurs.
If the contact never entered the workflow, use our separate guide to diagnose HubSpot workflow enrollment. If workflow errors reflect a wider handoff or reporting gap, you can discuss the process through our free Revenue Diagnostic.
HubSpot documentation
Troubleshoot common workflow errors; Understand the workflow details page and action logs; Workflows FAQ: action timing and throttling; Turn off workflows.
How did this article land?
Choose one reaction. You can change it anytime.