Zendesk Channel Setup

Zendesk setup in Jardine has two parts that must both be completed: account connection and inbound webhook configuration. Teams often finish the first part and assume they are done, then wonder why customer comments are not appearing in Conversations.

A reliable setup requires both pieces working together.

The Two-Part Model

Part one is OAuth connection. This allows Jardine to interact with your Zendesk environment and perform authenticated channel actions.

Part two is inbound event delivery. This is done through webhook URL configuration, trigger behavior in Zendesk, and signing secret setup in Jardine.

If either part is incomplete, behavior will look partial. You may have a connected badge but still miss live inbound traffic.

Step-by-Step Setup That Works

Open Settings -> Zendesk in Jardine.

Start by entering your Zendesk subdomain (the part before .zendesk.com) and run the connection flow. Once OAuth succeeds, status should show connected and the UI should display relevant connection details.

After connection, copy the org-specific webhook URL shown in the same tab. This URL is what Zendesk uses to send inbound comment events into Jardine.

Now move to Zendesk and create a webhook that points to that URL. Then create a trigger so customer-visible ticket comments are sent to that webhook. The trigger is what turns webhook configuration into actual event flow.

Finally, copy the webhook signing secret from Zendesk and save it in the Zendesk settings tab in Jardine. This step lets Jardine verify authenticity of inbound webhook payloads.

When these steps are complete, run a live test comment from a customer-side scenario and verify that the conversation appears and behaves correctly.

Why Signing Secret Setup Matters

Webhook URL alone is not sufficient for production safety. Signing secret verification ensures inbound payloads can be trusted as originating from your Zendesk configuration.

From an operations standpoint, this protects channel integrity. If secret handling is missing or mismatched, inbound processing may be rejected or treated as untrusted.

If you see connection status healthy but inbound behavior missing, secret mismatch is one of the first things to verify.

How to Test Before Broad Rollout

A good pre-launch test includes at least three comment types:

  • a routine support question,
  • a policy-sensitive question,
  • a high-friction customer message.

For each one, confirm arrival in Conversations, check status and ownership behavior, and confirm that escalation handling matches your policy expectations.

Do this with small controlled traffic first. Do not move to broad rollout until these checks are stable.

This process sounds conservative, but it is faster than debugging unpredictable behavior under full customer load.

Common Setup Mistakes

The first common mistake is subdomain mismatch. If the subdomain entered during connection is wrong, OAuth and webhook setup can become inconsistent.

The second mistake is skipping trigger setup in Zendesk. A configured webhook without a trigger often results in no inbound traffic.

The third mistake is forgetting to save or update the signing secret in Jardine after webhook changes.

The fourth mistake is testing with internal-only conditions that do not mirror real customer-visible comment flow.

The fifth mistake is assuming that a connected status badge means end-to-end channel readiness. It means connection readiness, not full operations readiness.

Operational Guidance After Launch

Once live, review Zendesk conversation samples regularly. Watch for patterns in escalations, ownership transitions, and unresolved thread clusters.

If behavior drifts, troubleshoot by layer:

  • ingress layer: webhook and trigger flow,
  • context layer: knowledge and connector coverage,
  • policy layer: routing and escalation expectations,
  • ownership layer: AI/human handoff correctness.

Layered diagnosis keeps fixes targeted and prevents random cross-setting changes.

Final Readiness Check

Zendesk channel setup is complete when your team can confidently say:

  • OAuth connection is healthy,
  • webhook URL and trigger are live,
  • signing secret is configured,
  • real customer comments enter Conversations,
  • handling outcomes match support policy.

When those conditions are true, Zendesk becomes a dependable production channel, not just an integration that “looks connected.”

After Zendesk is stable, continue with Routing Overview to shape escalation behavior more precisely, or Key Workflows to scale operations in controlled steps.