---
title: How to Import Tickets Without Sending Notifications (API Guide)
slug: how-to-import-tickets-without-sending-notifications-api-guide
date: 2026-11-17
author: Abdul Wahab
categories: [Migration Guide, Checklist, From The Migration Trenches]
excerpt: "Importing tickets into a live help desk can fire triggers, emails and assignment rules you cannot undo. Here is the import path, the pre-load checklist and the post-load checks that prevent it."
tldr: "Destination platforms treat created records as new work, so triggers, emails and webhooks can fire on years-old tickets. Use the import-specific endpoint where one exists, and switch off the rest before every load."
canonical: https://clonepartner.com/blog/how-to-import-tickets-without-sending-notifications-api-guide
---

# How to Import Tickets Without Sending Notifications (API Guide)


# How to Import Tickets Without Sending Notifications (API Guide)

Every record you create in a live help desk or CRM is a **business event**. The platform evaluates it against active triggers, automations, assignment rules and webhooks, then acts. It does not know that the ticket was closed in 2019 and is only arriving now because of a migration. The result can be acknowledgement emails to customers about old issues, historical cases routed into today's queues, and downstream systems ingesting history as new activity.

Deleting the records does not recall the emails.

To import tickets without sending notifications, use the platform's historical import endpoint where one exists, such as Zendesk's Ticket Import API, and switch off the triggers, rules and webhooks that could still react. This guide covers both, with the exact behaviour Zendesk and Salesforce document, and the sequence to pause and restore a live system's rules.

*Last verified: 17 November 2026 against Zendesk and Salesforce documentation.*

## Why destination platforms fire on imported records

**A created record is a created record.** A help desk's rules engine evaluates each new record against its active rules and runs the matching actions. That is what makes automations dependable in daily operations. During a migration, the same behaviour turns history into fresh work.

Zendesk says it plainly. Every ticket created with the Create Many Tickets endpoint "may be affected by your business rules, which can include sending email notifications to your end users." For historical tickets, or more than 1,000 tickets, Zendesk points to the Ticket Bulk Import endpoint instead ([developer.zendesk.com](https://developer.zendesk.com/api-reference/ticketing/tickets/tickets/)).

This is where a general-purpose agent working from public documentation tends to go wrong. Vendor MCP servers are built for conversational work on a few records at a time, and their create tools call the normal create path. The import endpoint, the order of operations and the list of rules to pause are migration knowledge. They are not something a create tool supplies.

A **delta sync**, a follow-up pass that moves only the records that changed after the bulk load began, opens a second risk window. A delta often updates tickets that were already imported. On Zendesk, an update to an imported ticket brings triggers back into play, so the delta needs the same controls as the bulk load.

For the wider picture of agent-run migrations, see [AI agent data migration via MCP: the 2026 guide](https://clonepartner.com/blog/blog/ai-agent-data-migration-via-mcp-the-2026-guide). For moving the automations themselves, read our guide on [how to migrate automations, macros and workflows](https://clonepartner.com/blog/blog/how-to-migrate-automations-macros-workflows).

## What Zendesk does, and the endpoint that prevents it

The **Ticket Import API** is Zendesk's path for historical tickets. It has two endpoints: `POST /api/v2/imports/tickets` for one ticket and `POST /api/v2/imports/tickets/create_many` for up to 100 tickets per request. Both are allowed for admins only ([developer.zendesk.com](https://developer.zendesk.com/api-reference/ticketing/tickets/ticket_import/)).

Zendesk documents this behaviour for imported tickets ([developer.zendesk.com](https://developer.zendesk.com/api-reference/ticketing/tickets/ticket_import/)):

- "If a ticket with a status other than closed is imported, triggers won't run on the imported ticket. However, if the ticket is updated after import, triggers will resume."
- No notifications are sent to users CC'd on imported tickets. Notifications are sent on subsequent updates.
- Zendesk metrics and SLAs are not supported for imported tickets. Running them on imported tickets gives incomplete and inaccurate data.
- Zendesk recommends adding a tag to imported tickets and excluding them from reports.

The catch is in the first bullet. **The import suppresses triggers at import time only.** Any later update, from a delta sync, a field fix-up or a bulk tag change, can fire triggers and send notifications unless you have paused them.

The two bulk endpoints look alike in code. They behave differently:

| Behaviour | Create Many Tickets<br>`POST /api/v2/tickets/create_many` | Ticket Bulk Import<br>`POST /api/v2/imports/tickets/create_many` |
|---|---|---|
| Business rules on create | "May be affected by your business rules", including email notifications | Triggers won't run if the imported status is anything other than closed |
| CC notifications on create | Subject to your business rules | None on import, sent on subsequent updates |
| `created_at` | Read-only, so the ticket gets the current time | Can be set, with `updated_at` and `solved_at`, plus comment `created_at` |
| Metrics and SLAs | Standard behaviour | Not supported for imported tickets |
| Tickets per request | Up to 100 | Up to 100 |
| Allowed for | Agents | Admins only |

Sources: [Tickets API](https://developer.zendesk.com/api-reference/ticketing/tickets/tickets/) and [Ticket Import API](https://developer.zendesk.com/api-reference/ticketing/tickets/ticket_import/).

### Should you disable all Zendesk triggers?

Usually not. Zendesk's default triggers include "Notify requester and CCs of received request" and "Notify requester and CCs of comment update" ([support.zendesk.com](https://support.zendesk.com/hc/en-us/articles/4408828984346)). Deactivating every trigger stops live customer emails too.

The better pattern is to use the Import API for the historical load and pause only the triggers that could react to follow-up updates or a delta sync. Record each trigger's state before you touch anything, so you can restore exactly what was there.

### Test outside production first

Run the full procedure in a Zendesk sandbox or a test account before you touch production. Check that the import endpoint behaves as documented for your statuses, that batching and retries work, and that the email log shows nothing unexpected. Use test requester addresses you control, so a misfire reaches your team rather than customers.

### Rate limits, timestamps and archived tickets

The bulk import endpoint accepts up to 100 tickets per request ([developer.zendesk.com](https://developer.zendesk.com/api-reference/ticketing/tickets/ticket_import/)). Requests count against the Support API rate limit for your plan: 200 requests per minute on Team, 400 on Growth and Professional, 700 on Enterprise and 2,500 on Enterprise Plus or with the High Volume API add-on ([developer.zendesk.com](https://developer.zendesk.com/api-reference/introduction/rate-limits/)). Build retries with backoff.

The Import API lets you set `solved_at`, `updated_at` and `created_at` on tickets, and `created_at` on comments. A ticket `created_at` or `updated_at` before 1970 is rounded up to 1970. A comment `created_at` cannot be before 1970 or in the future ([developer.zendesk.com](https://developer.zendesk.com/api-reference/ticketing/tickets/ticket_import/)). For the full field-by-field picture, see [how to keep the created date when migrating](https://clonepartner.com/blog/blog/keep-created-date-when-migrating-system-fields-import-apis).

Ticket imports also accept an `archive_immediately` parameter. With it, a ticket imported with a `closed` status goes straight to the ticket archive. Zendesk suggests exploring it for large historical imports of 750,000 tickets or more ([developer.zendesk.com](https://developer.zendesk.com/api-reference/ticketing/tickets/ticket_import/)).

> **Archived tickets need a different export.** "Archived tickets are not included" in List Tickets responses. To get every ticket, Zendesk points to the Incremental Ticket Export ([developer.zendesk.com](https://developer.zendesk.com/api-reference/ticketing/tickets/tickets/)), which is limited to 10 requests per minute, or 30 with the High Volume add-on ([developer.zendesk.com](https://developer.zendesk.com/api-reference/introduction/rate-limits/)).

## How to disable workflow rules before a Salesforce data load

Salesforce's guidance for large loads is to switch rules off around the load, then switch them back on. Its Trailhead module on loading large data volumes says you can "disable data validation and enrichment operations while loading" without compromising data integrity or business rules. Afterwards, "Turn validation, workflow, and assignment rules back on" so they act as users enter and edit records ([trailhead.salesforce.com](https://trailhead.salesforce.com/content/learn/modules/large-data-volumes/load-your-data)).

### Set rules to inactive

The same module says to edit each validation rule and set it to inactive. You disable lead and case assignment rules and territory assignment rules the same way. It also warns that workflow rules "can also slow inserts to a crawl" if they are on during a massive load ([trailhead.salesforce.com](https://trailhead.salesforce.com/content/learn/modules/large-data-volumes/load-your-data)).

Before the load, deactivate:

- Workflow rules on the objects you are loading
- Lead and case assignment rules
- Territory assignment rules
- Validation rules that legacy data would fail

### Bypass Apex triggers with a custom setting

Triggers take more preparation. Trailhead describes creating a custom setting with a checkbox field that controls when a trigger fires, and adding a statement to the trigger code that checks it. After that, "disabling or enabling your trigger is as simple as editing the checkbox field" ([trailhead.salesforce.com](https://trailhead.salesforce.com/content/learn/modules/large-data-volumes/load-your-data)).

An illustration of that check, with example names:

```apex
// In your trigger or trigger handler
Migration_Settings__c settings = Migration_Settings__c.getInstance();
if (settings != null && settings.Bypass_Triggers__c) {
    return; // Skip trigger logic while the checkbox is ticked
}
```

After the load, reset the checkbox so triggers fire normally again. Our deep dive on [managing Apex triggers during bulk Salesforce data loads](https://clonepartner.com/blog/blog/how-to-manage-apex-triggers-during-bulk-salesforce-data-loads) covers the full pattern.

### Clear the Data Loader assignment rule

Data Loader has an **Assignment rule** setting. It takes the ID of an assignment rule to use for inserts, updates and upserts on cases and leads, and that rule overrides the Owner values in your file ([developer.salesforce.com](https://developer.salesforce.com/docs/platform/data-loader-impl/guide/configuring-the-data-loader.html)). Leave it empty for a historical load, or your file's owners are replaced and old cases are routed as new work.

### Enable audit fields for historical dates

To keep original dates, you set `CreatedDate`, `CreatedById`, `LastModifiedDate` and `LastModifiedById` on insert. Salesforce's setup is: from Setup, open **User Interface**, enable the "Set Audit Fields upon Record Creation" and "Update Records with Inactive Owners" user permissions, then grant "Set Audit Fields upon Record Creation" in the permission set or profile that will load the data. `SystemModstamp` is the one audit field you cannot set ([developer.salesforce.com](https://developer.salesforce.com/docs/platform/object-reference/guide/system-fields.html)).

## What goes wrong: five failure modes

Each of these follows directly from the platform behaviour above.

### 1. Customers receive "we got your request" emails for old tickets

A script creates 15,000 historical tickets through Zendesk's standard create endpoint. The default trigger "Notify requester and CCs of received request" emails the requester and everyone copied ([support.zendesk.com](https://support.zendesk.com/hc/en-us/articles/4408828984346)). Customers who last wrote in years ago get a fresh acknowledgement. Some reply, which opens new tickets.

### 2. Reports count history as new work

A ticket created through the standard endpoint carries the current time, because `created_at` is read-only there ([developer.zendesk.com](https://developer.zendesk.com/api-reference/ticketing/tickets/tickets/)). Years of history land in this week's volume charts. Tickets created through the Import API keep their dates, but Zendesk says metrics and SLAs are not supported for them, so they still need a tag and an exclusion from reports ([developer.zendesk.com](https://developer.zendesk.com/api-reference/ticketing/tickets/ticket_import/)).

### 3. Assignment rules flood agent queues

A Salesforce case load runs with an assignment rule set in Data Loader. The rule overrides the owners in the file ([developer.salesforce.com](https://developer.salesforce.com/docs/platform/data-loader-impl/guide/configuring-the-data-loader.html)), and thousands of old cases are routed into today's queues. Agents start ignoring the queue, and real cases get missed among the imports.

### 4. Webhooks push imported data downstream

A help desk has a webhook that posts each new ticket to a Slack channel and an analytics service. Every imported ticket fires it. A load of 5,000 tickets means 5,000 events. The channel fills with noise, and the analytics service counts history as new activity.

### 5. Deleting the records does not undo the damage

The instinct after a mass-notification incident is to delete the imported records and start again. The emails have already been sent. The webhook payloads have already been delivered. Downstream systems have already ingested the data. The records go. The side effects stay.

> **You cannot recall an email by deleting the ticket that triggered it.** Once an automation fires, its side effects are permanent. Prevention is the only fix.

## Partial recovery after a mass-notification incident

If notifications have already gone out, this sequence limits further damage and gives you a record for stakeholders.

**Immediate actions:**
1. Stop the import. Do not keep loading while automations are active.
2. Deactivate the specific triggers that fired. Leave the others on, so live customers still get their emails.
3. Check the email logs to size the problem: how many notifications, to how many recipients, over what time window.

**Customer communication:**
4. Send affected customers a short message from a named person, not an automated queue. Say they received a notification in error and need to do nothing. This keeps them from opening new tickets in reply.
5. If affected customers have account managers, tell the managers first.

**Downstream systems:**
6. Contact the owners of analytics and BI systems. They may be able to flag or delete the records before they reach reports.
7. For webhook payloads already delivered, check whether the receiving system processed the same record twice. A deduplication pass needs a unique key, such as the source ticket ID, on every record.

**Root cause:**
8. Write down which endpoint or tool caused the incident, which automations fired, and what the corrected procedure is.

## Pre-load checklist: what to disable before importing

A **pre-load checklist** is a control list of every destination feature that can treat imported records as new work. Walk through it before the first record reaches the destination.

### Triggers and automations

- **Zendesk:** Use the Ticket Import API, not the standard create endpoint. Test the procedure outside production first. Pause the notification triggers that could fire on post-import updates or the delta sync. Record each trigger's state before you change it.
- **Salesforce:** Set validation rules, lead and case assignment rules and territory assignment rules to inactive. Switch off workflow rules on the loaded objects. Tick the custom setting checkbox that bypasses triggers. Confirm the loading user has "Set Audit Fields upon Record Creation".
- **Any other destination:** Find the vendor's own documentation on importing history. If it does not describe a path that suppresses notifications, assume normal business rules apply and switch off the rules that match created records.

### SLA policies

- Decide how imported tickets relate to SLA reporting before the load. On Zendesk, metrics and SLAs are not supported for imported tickets, so tag them and exclude the tag from reports.
- If reporting continuity matters, carry elapsed time and original breach status into custom fields. They cannot be rebuilt after the load.

### Assignment and routing rules

- **Salesforce:** Deactivate lead and case assignment rules, and leave the Data Loader Assignment rule setting empty.
- **Zendesk:** Triggers that reassign on group, priority or tags can run once an imported ticket is updated. Pause them for the delta sync and any fix-up updates.

### Webhooks and integrations

- Pause outbound webhooks, Slack integrations, automation-platform workflows and any middleware that listens for record-creation events.
- If you cannot pause a webhook, make sure every payload carries a unique key, such as the source ticket ID, so the receiving system can detect duplicates. Pausing is safer than relying on deduplication.
- Check for native integrations that react to new records independently of outbound webhooks.

### Email channels

- If the destination has no import path that suppresses notifications, switch off the notification rules for the import window. Tell your support team that live tickets filed in that window may need manual follow-up, and record the exact start and end times.

## Post-load verification: how to prove nothing fired

Restoring automations without checking is how a second wave of incidents starts. *"No one complained"* is not a control. Verify with system evidence.

### Before re-enabling anything

- **Check email logs.** Look for any outbound email in the import window that you did not expect.
- **Check ticket event history.** Open a sample of imported tickets, including the oldest, the newest and some open ones, and read their events for trigger activity. Check webhook and integration logs as well, rather than trusting one surface.
- **Check webhook delivery logs.** For each webhook that fired, confirm whether the receiving system processed the payload and whether it needs deduplication.
- **Check downstream systems.** Look at Slack channels, analytics platforms and any system fed by your help desk or CRM for imported records that should not be there.

### Re-enable automations in order

Re-enable in this order, so a misconfigured rule meets a test record before it meets live traffic:

1. **Assignment rules first.** Confirm a live test ticket routes to the correct queue or owner.
2. **Triggers and automations second.** Create one test ticket and confirm the expected notifications fire and no others.
3. **Webhooks and integrations last.** Send a test event and confirm the receiving system processes it once.

### Run a controlled delta sync

The delta sync moves only the records that changed after the bulk load began. On Zendesk, it updates tickets that were already imported, and triggers resume on an updated imported ticket. Keep the notification triggers paused for the delta, and run the same checks on the delta batch.

> **Tag everything.** Add a tag such as `migrated` to every imported record. It lets you exclude history from reports, filter it out of trigger conditions, find it in downstream systems and scope a recovery if something goes wrong. Zendesk recommends exactly this for imported tickets. Set the tag in the import call itself, not in a later bulk update, because on Zendesk that update would bring triggers back.

## Where AI genuinely helps with this

AI is useful around the load. It can turn an exported list of triggers and automations into a draft pause-and-restore checklist. It can write the loader code and the tests that check a sample for side effects, and that code then runs outside the model. It can also review a sample's ticket events and flag anything that looks like a trigger firing.

The pattern that works: AI drafts, a person reviews, and deterministic code runs the job. The help desk version of this argument is in [can an AI agent migrate your help desk?](https://clonepartner.com/blog/blog/can-an-ai-agent-migrate-your-help-desk-zendesk-intercom-freshdesk).

## The risk is not data loss, it is data side effects

Import endpoints and loaders handle field values well. The bigger risk is that the destination treats imported records as live business events and acts on them before you can intervene. The fix is procedural: inventory the destination's automations, pause them, test outside production, import, verify, restore.

Skip that process, and you are one API call away from emailing every customer in your database about a ticket they filed three years ago.

> Planning a help desk or CRM migration into a live system? Book a free 30-minute scoping call. A migration engineer will map which triggers, rules and webhooks to pause, and you get a fixed-price proposal with no hourly billing.
>
> [Talk to us](https://clonepartner.com/talk-to-us?duration=30&utm_source=blog&utm_medium=button&utm_campaign=demo_bookings&utm_content=cta_click&utm_term=demo_button_click)

## Frequently asked questions

### How do I import tickets into Zendesk without sending notifications?

Use Zendesk's Ticket Import API (POST /api/v2/imports/tickets, or /api/v2/imports/tickets/create_many for up to 100 tickets per request), not the standard Create Many Tickets endpoint. Zendesk says triggers won't run on an imported ticket whose status is anything other than closed, and no notifications go to users CC'd on imported tickets. If the ticket is updated after import, triggers resume and notifications are sent on subsequent updates, so plan the follow-up updates and delta sync with that in mind.

### How do I disable workflow rules before a Salesforce data load?

Salesforce's Trailhead module on loading large data volumes says to edit validation rules, lead and case assignment rules, and territory assignment rules and set them to inactive. For triggers, it describes a custom setting with a checkbox field that the trigger code checks. Afterwards, turn validation, workflow and assignment rules back on and reset the custom setting.

### Will imported tickets show up in Zendesk SLA and metrics reports?

Zendesk says metrics and SLAs are not supported for tickets created through the Ticket Import API, and that running them on imported tickets gives incomplete and inaccurate data. Zendesk recommends adding a tag to imported tickets and excluding them from reports.

### Can I undo emails sent by automations during a data import?

No. Deleting the imported records does not recall emails that were already sent or reverse webhook deliveries that were already made. The only reliable fix is prevention: use the import path and switch off automations before the load.
