Zendesk → Pylon

Catalog plan outline

General guidance from the Afterstack catalog for Zendesk → Pylon. Not Zendesk or Pylon vendor documentation.

Migration friendliness

Zendesk 7/10 · Pylon 5/10

Speed to live

Zendesk 5/10 · Pylon 7/10

Data to export from Zendesk

• custom objects
• call and voice logs
• synced email and calendar links
• workflow rules
• saved reports
• installed marketplace apps
• AI assistant configuration
• records for the catalog use cases: accounts

How the catalogs differ

• Zendesk is flagged for custom objects. Pylon is not, so the map is a redesign rather than a field-for-field copy.
• Pylon is not flagged for: custom objects, call and voice logs, installed marketplace apps.
• Zendesk is flagged with an open API, which is the export path to test.
• Pylon is flagged with an open API for the load.
• Zendesk is cataloged as a fit for: Multi-channel support orgs that need SLAs, QA, and a ticket system of record.
• Pylon is cataloged as a fit for: B2B SaaS support.

Phases

1. Export

   Take Zendesk as the source of truth for this pair. Export custom objects; call and voice logs; synced email and calendar links; workflow rules; saved reports; installed marketplace apps; AI assistant configuration; records for the catalog use cases: accounts. The default enterprise helpdesk — tickets, SLAs, and a marketplace around them.

2. Map

   Load into Pylon only after the field map is written down. B2B support hub for Slack, Teams, and ticket-weary account teams. Zendesk is flagged for custom objects. Pylon is not, so the map is a redesign rather than a field-for-field copy.

3. Prove a sample

   Pylon scores 7/10 on speed to live and 5/10 on migration friendliness. Run one sample load and one report the team already uses in Zendesk before a full cutover.

4. Cut over

   Zendesk scores 7/10 on migration friendliness. Keep Zendesk readable until the team is working in Pylon. Zendesk is the procurement-safe customer service platform: ticketing, messaging, QA, and a large app ecosystem. AI Copilot and the Sunshine layer sit on top of a ticket model most support orgs already know. It is powerful, configurable, and easy to outgrow in cost once every add-on is on.

Zendesk to Pylon: what the vendor docs actually cover

Linked sentences come from vendor documentation opened for this page. Everything else is general guidance, not Zendesk or Pylon vendor documentation. This section does not move data and does not change the catalog outline above.

What to export from Zendesk

Zendesk's export article is the only source-side document this cutover uses. Exports are switched off until the account owner contacts Zendesk Support and includes the subdomain. JSON, CSV, and XML tools are not offered on Team. The REST API is. The admin path, once exports exist, is Admin Center, Account, Tools, Reports. JSON is the format the article recommends once a ticket archive passes about two hundred thousand tickets, and that JSON download is not available in a sandbox. CSV leaves out comments, leaves out descriptions, leaves out deleted tickets, and drops any ticket bigger than one megabyte. JSON is NDJSON. It skips items whose timestamp is within six minutes of the request. A ticket over one megabyte omits comments from the main file and adds a second file that explains the omission. Tickets that belong to AI agents are outside the export. XML is described around a five hundred megabyte or two hundred thousand ticket ceiling, and it does not carry custom ticket fields. The download link is valid for at least three days. Order is not guaranteed. Organization JSON can include deleted organizations, marked with deleted_at. Secondary emails are not in the main user file. The article points those at the List Users API with include=identities. This pass did not re-fetch Zendesk's pricing page, so no new Zendesk seat price is stated here.

• Zendesk export tickets, users, or organizations — checked 2026-10-12

How Pylon imports

Pylon's pricing path, requested on 2026-10-12, redirects to a page that schedules a demo. No seat dollar is printed there. The catalog's custom model and null starting price match that absence. The catalog note about SOC 2 Type 2 and a November 2022 founding date was not re-checked on a security page during this pass, so those sentences are not repeated as fresh findings. The docs introduction at docs.usepylon.com describes an agentic support product, connectors, an API entry in the navigation, automations, and skills. It does not describe a Zendesk ticket importer, a CSV mapper, or a place to put NDJSON. A docs-search URL is a search box, not a migration guide, and it is not cited. Every object match below is general guidance beside those two Pylon pages.

• Pylon pricing — checked 2026-10-12
• Pylon docs introduction — checked 2026-10-12

Object and field mapping

• Zendesk tickets are the conceptual source for a support conversation. The export article can produce them as JSON, CSV, or XML after Support enables the tools. Pylon's docs intro does not name a ticket file.
• Zendesk users are the conceptual source for requesters and agents. Secondary emails require a separate API call. Do not assume the CSV user file is complete.
• Zendesk organizations can travel in JSON, including rows with deleted_at. A destination that rejects deleted rows needs a filter the Zendesk article does not apply for you.
• Comments and descriptions exist in JSON, with the one-megabyte exception, and do not exist in CSV. Pick the format before you promise agents their history.
• Custom ticket fields are absent from XML. If those fields are the business, XML is the wrong file.
• AI-agent tickets are excluded. Map them as a known hole, not as a Pylon failure.

What will not carry over

• The admin export on Team plans. The API remains.
• AI-agent tickets.
• CSV comments, CSV descriptions, and deleted tickets.
• Comments on JSON tickets larger than one megabyte, except the explanatory side file.
• Custom ticket fields inside XML.
• A guaranteed row order, or a download link older than the article's three-day minimum.
• A Pylon importer. None was documented on the pages opened.

Cutover checklist

1. Have the account owner ask Zendesk Support to enable exports, and include the subdomain.
2. If the plan is Team, plan an API pull instead of Admin Center.
3. Choose JSON once the archive is large. Do not run that JSON job in a sandbox.
4. Reject CSV if agents need comment text.
5. Split or specially handle tickets over one megabyte.
6. Pull secondary emails with List Users and include=identities.
7. Download before the link ages past three days, and do not sort-depend on the file.
8. Keep Zendesk open, because Pylon's docs do not say the file has a landing zone.

Common pitfalls

• Citing a docs search URL as if it were an import specification.
• Treating a deleted organization in the JSON as a live account. Read deleted_at.
• Promising a Monday full download while exports are still disabled.
• Using XML and then wondering where custom fields went.
• Inventing a Pylon price because the demo page has none. The catalog already says there is no public list price.

Related catalog pages

• Best AI-native Zendesk alternatives (2026) — Which tool, compared
• What should replace Zendesk? — Switch narrative
• Pylon product page — B2B support hub for Slack, Teams, and ticket-weary account teams.
• Pylon vs Zendesk — Head-to-head already in the catalog
• Customer support category — Catalog hub for this category