Skip to content

Customer support

How to find and size your Zendesk data.

Find how many tickets Zendesk Support holds, archived ones included, and how a full-history JSON export or the incremental export API works.

Find how much you have

  1. Decide what you are sizing. This guide covers Zendesk Support tickets. Help center articles, messaging and live chat conversations, and Explore reports each have their own export route, so size them separately only if they are in scope. If you use an AI agent, note roughly how many conversations it handled without a human, because Zendesk treats those AI agent tickets differently.

  2. Get a total that includes archived tickets. Zendesk archives a ticket 120 days after it is closed in most cases, and archived tickets drop out of views and out of some API endpoints. Zendesk's archiving page lists search as a way to reach them, so a developer should call GET /api/v2/search/count with the query type:ticket. The search API lists at most 1,000 results per query, but Zendesk says the count it reports is the actual number of matches.

  3. Split the count by year. Add a date filter to the query, such as type:ticket created<2022-01-01, or a range with created> and created< together. Zendesk's search takes dates as YYYY-MM-DD. Run it once per year. The yearly counts should add up to the total, and a year that returns zero tells you where your history begins.

  4. No developer? Ask the account owner to enable exports, run a JSON export over a wide date range, and count the tickets. Zendesk writes the export as newline-delimited JSON, which puts one record on each line, so the line count across the ticket files is your ticket count. Zendesk says tickets over 1 MB are still listed in the main file, only without their comments, and are listed again in a second file with an error message, so leave that second file out of the count.

  5. Cross-check with GET /api/v2/tickets/count. Zendesk calls that figure approximate and says it refreshes only once a day when the count passes 100,000. Its archiving page also lists Count Tickets among the endpoints that skip archived tickets, so expect a lower number than the search count.

  6. Find the start date by searching with type:ticket order_by:created sort:asc and reading the created date of the first result. If you moved from another help desk, say so in your description.

What to put on your intake

Your intake has five boxes for each system. Here is what goes in each one for Zendesk.

What does it hold?
Support tickets with every comment, tags, custom fields and audit events, from the channels you use (email, web form, chat, phone). Say which brands or help desks are in scope, and note roughly how many conversations an AI agent handled without a human.
Data types
Text for ticket threads and Structured records for fields, tags and status history. Add Image only if you plan to include attachments.
How big is it?
Tickets, from the search count that includes archived tickets. Put the number of agents or brands and the share of tickets with attachments in the description if you know them.
Verified or estimated
Verified when the number is the Search API count, or a line count of a full JSON export. Estimated when it comes from a view, from the Count Tickets endpoint, or from one year scaled up.
Records start
The created date of the oldest ticket, from the ascending search above. Mention a migration if early tickets were imported from another tool.

Exporting the full history

If a buyer licenses your records, this is how the full history comes out of Zendesk.

  1. Nothing is exported until a buyer licenses the data and the scope is agreed: which brands, which years, and whether attachments and user records are in. The steps below show what to expect.

  2. Exports are off by default. The account owner contacts Zendesk Customer Support to enable them. They are available on Support Professional and Enterprise and on Suite Growth, Professional, Enterprise and Enterprise Plus. Team plans have no export tool, but every plan can use the API. The approved email domain for exporting admins starts out as the account owner's domain.

  3. In Admin Center, go to Account, then Tools, then Reports, and open the Export tab. Choose the JSON export, which Zendesk recommends for accounts with more than 200,000 tickets, pick tickets, users or organizations plus a date range, and click Export. The date range uses a system-generated timestamp that typically matches a ticket's latest update, so choose a wide range, not creation dates. Zendesk emails a download link, valid for at least three days. It can take a few minutes or a day or more.

  4. The JSON comes as newline-delimited records. Accounts with more than a million tickets are downloaded in 31-day increments. A ticket over 1 MB is exported without its comments, and a second file lists those tickets with an error message. The CSV export is no substitute: it covers tickets only and leaves out comments and descriptions. The full XML export is not available above 200,000 tickets.

  5. A developer can pull data through the incremental export API instead. GET /api/v2/incremental/tickets/cursor.json with a start_time returns tickets created or changed on or after that time. Follow each page's cursor until end_of_stream is true. Ticket events only support the time-based endpoints. Zendesk limits incremental exports to 10 requests a minute, or 30 with the High Volume API add-on, and each page holds up to about 1,000 items. Run it from your own account, because Zendesk restricts third-party integrations and Marketplace apps from these endpoints.

  6. Under the license, names, email addresses, phone numbers and account or order numbers of customers and agents are removed from ticket text, subjects and fields before delivery. Attachments are excluded unless a specific package is agreed.

We only ask for an export after you sign a license and agree the scope. Sizing your data for the intake never needs one.

Limits to know

  • A missing Export tab usually means exports were never enabled, your plan does not include them, or your email domain is not on the approved list. The fix is usually the account owner contacting Zendesk Customer Support, or an admin checking permissions and the approved domain. JSON and CSV exports are not available in sandbox accounts.
  • Zendesk says AI agent tickets, meaning conversations an AI agent handled without being escalated to a human, cannot be exported with the data export tools, and ticket-listing APIs leave them out by default. A developer can request them from the tickets and incremental tickets endpoints by adding support_type_scope=ai_agent. Once a conversation is escalated it becomes a regular ticket.
  • Archived tickets are included in search, the incremental export and the full export, but they are left out of views and of several API endpoints, including Count Tickets and Listing Tickets. Tickets with more than 10,000 events or more than 2 MB of associated records are not archived, and very high-volume accounts may see archiving sooner than 120 days.
  • When a ticket is archived, its whole event history moves to the time of the close. Zendesk's incremental export page warns that a request starting in 2019 can return events stamped in 2014, so do not read event times as a clean timeline.
  • Deleted tickets sit in the Deleted tickets view for 30 days and are then permanently deleted, with only the ticket ID kept and all user-submitted content scrubbed. Scrubbed versions still appear in incremental export results, and Zendesk removes them from exports 90 days after permanent deletion.
  • Zendesk's JSON export drops anything changed within 6 minutes of the request, and the CSV and XML exports have their own limits. A full XML file tops out at 500 MB, roughly 200,000 tickets, so do not rely on it for a large account.
  • The export pages do not say whether attachment files travel with the data. Ask Zendesk before you promise a buyer screenshots or documents.
  • Canceling does not leave your data in place for long. Zendesk's policy starts deleting data 90 days after cancellation, and it deletes ticket data, user data and attachments within 40 days after that process begins. To export after canceling, ask Zendesk Customer Support for a temporary reactivation within those 90 days.
  • Support API calls count against a per-minute limit that depends on your plan, from 200 requests a minute on Team to 700 on Enterprise and 2,500 on Enterprise Plus. Going over returns HTTP 429, and the Retry-After header says how many seconds to wait.

This guide isn't legal advice. Check your contracts, and any recording-consent rules that apply, with your own lawyer.

Why buyers value it

Support tickets are the clearest record of what customers actually get stuck on and how a company's people fix it. A ticket holds the problem in the customer's own words, each reply an agent wrote, the tags and fields that routed it, and how it ended. Buyers building support agents want whole threads with outcomes, including the policies agents had to follow and the moment a case was escalated.

A ticket is worth more when it connects to the engineering work that fixed its cause. Years of tickets from one company also show which problems repeat and how the answers changed as the product did.

What stays private

Buyers receive de-identified operating records. Under the license, the de-identification step before delivery removes or excludes:

  • People: requester and agent names, email addresses, phone numbers and usernames, in ticket text, subjects, tags and custom fields.
  • Customer details: company names, order, invoice and account numbers, and addresses that appear in ticket text.
  • Sensitive content: passwords, keys, payment details and health information that customers or agents typed into a ticket. Tickets from customers with payment or health data can be left out of scope altogether.
  • Attachments: screenshots, invoices and identity documents, unless a specific package is agreed.
  • Internal notes: private comments that hold credentials or personal remarks about customers.

Common questions

Exports are off until the account owner asks Zendesk Customer Support to enable them, and they depend on your plan. Support Professional, Support Enterprise and Suite Growth and above include them, and Team plans do not. Only admins and custom roles with export permission can run one, and your email domain must match the approved domain. If you cannot export, a developer can use the API on any plan.

Yes, they are still your tickets. Zendesk archives closed tickets, usually 120 days after they close, and keeps them in search, in the incremental export and in the full export. They disappear from views and from the Count Tickets endpoint, which is why those numbers run low. Use a search count or a full export for your intake figure.

Not through the export tools. Zendesk describes AI agent tickets as conversations handled entirely by an AI agent or a default messaging response, never escalated to a human, and says they cannot be exported. Ticket-listing APIs skip them by default, but a developer can pull them from the tickets and incremental tickets endpoints with support_type_scope=ai_agent. Once a conversation is escalated, it becomes a regular ticket. Mention the AI agent's share of your volume in your description so the picture stays honest.

Sources

Vendor documentation, checked . Screens and plans change, so if a step looks different in your account, check with Zendesk.