Integrations & Auth 8 min read

How to know if a Calendly OAuth token has expired

Fast diagnostic summary

A Calendly setup depends on two separate OAuth token chains: the upstream credential linking Calendly to each host's Google Workspace or Microsoft 365 mailbox, and any downstream OAuth tokens linking Calendly's v2 API to your CRM or availability monitor. When an upstream calendar refresh token is revoked by a password change or Conditional Access policy, Calendly usually fails closed and hides every bookable slot while emailing only that individual host — leaving firm leadership unaware until prospects report an empty page.

Two OAuth layers, two very different failure signatures

When an operations manager hears "the Calendly OAuth token expired," they need to isolate which handshake broke before telling staff to reconnect accounts blindly:

  • Layer A — Upstream Calendar Sync (Calendly ↔ Google / Exchange / iCloud): Each host authorizes Calendly to read free/busy windows and write calendar invites. When this refresh token dies, Calendly can no longer confirm whether Tuesday at 2:00 PM is free.
  • Layer B — Downstream App OAuth (Calendly API ↔ HubSpot / Salesforce / AvailTrack): Organization admins or users authorize external tools via Calendly's OAuth 2.0 server (auth.calendly.com/oauth/token), where short-lived access tokens rotate alongside single-use refresh tokens.

Layer A failures stop clients from booking immediately. Layer B failures leave the public booking page live while breaking lead routing, webhook registration, or automated oversight — unless your monitoring platform explicitly alerts on refresh-token failures as an incident class.

Enterprise IT events that silently kill upstream calendar tokens

OAuth 2.0 refresh tokens are designed to renew silently for months, which lulls regulated wealth advisory and consulting firms into assuming calendar connections are permanent. In corporate Google Workspace and Microsoft Entra ID tenants, six routine IT events invalidate refresh grants immediately:

Identity / IT trigger Affected stack Why the token dies Public booking symptom
Conditional Access Sign-In Frequency cap (e.g., 30 or 90 days) Microsoft 365 / Entra ID Refresh token hits tenant max-age policy requiring interactive MFA 0 slots shown on host's links
Password reset or compromised-session "Revoke all sessions" M365 & Google Workspace Security token version increments; Graph & Google Calendar scopes invalidated Immediate fail-closed blackout
Google Cloud OAuth app restricted to "Trusted" allowlist Google Workspace Workspace admin blocks unverified or non-allowlisted Calendar API scopes Org-wide simultaneous disconnect
Primary UPN / email domain rename after merger or rebrand M365 & Google Workspace Mailbox principal identifier changes beneath existing calendar binding Conflict check errors / failed writes
Apple ID app-specific password rotation iCloud CalDAV Partner changes main Apple ID password, wiping app-specific CalDAV secrets Personal conflict calendar drops
Shared sub-calendar ACL downgraded or deleted Exchange / Google Host loses freeBusyReader access to a checked secondary calendar Either 0-slot block or stale sync

Fail-closed vs. fail-open: why token expiry hides every slot

When Calendly cannot reach a host's primary conflict calendar because Microsoft Graph returned 401 InvalidAuthenticationToken or Google returned invalid_grant, Calendly faces a design choice: guess that the advisor is free (risking embarrassing double-bookings over client trials or board calls), or assume no slot can be safely verified.

For primary conflict calendars, Calendly sensibly chooses safety and fails closed — suppressing every candidate slot until the calendar connection is re-established. To an outside prospect, an expired OAuth token looks indistinguishable from a partner who is 100% booked through next month (one of the hardest-to-diagnose variants of Calendly showing no availability).

Even worse is why the outage lingers: Calendly sends the "Reconnect your calendar" email to the individual host's inbox. Busy partners receive dozens of automated vendor emails daily, archive the warning assuming IT will handle it, and never open calendly.com/app/availability/calendar_settings because their existing Outlook appointments keep appearing normally.

Four-step diagnostic runbook to confirm an expired token

If a practitioner's booking page abruptly goes dark despite open space on their calendar, run these four checks in order — total triage takes under three minutes:

Step 1: Check Calendly Troubleshooting codes on the live event type

Logged in as the host (or viewing via Admin troubleshooting), click Troubleshoot at the bottom of the live booking calendar and select a date that should be open. Rather than guessing, inspect the exact block reason tag: CALENDAR_ERROR or conflict-fetch warnings point directly to a broken calendar integration, whereas NOCAP, BUFFER, or OUTSIDE_HOURS exonerate OAuth immediately.

Step 2: Inspect Calendar Sync status & sub-calendar checkboxes

Open Account → Calendar Sync. A disconnected Google or Office 365 credential displays a red reconnect banner. After clicking Reconnect, always verify two settings that frequently reset during re-authentication: which sub-calendars are checked under Check for conflicts, and which target calendar receives new bookings under Add to calendar.

Step 3: Audit Microsoft Entra / Google Workspace sign-in logs

If multiple staff members lose sync every 30 days on the dot, ask IT to inspect Entra ID Conditional Access policies affecting third-party OAuth enterprise applications. Exempting verified scheduling apps from aggressive browser-style session timeouts — or migrating from legacy Exchange EWS connectors to modern Graph OAuth — stops recurring mass disconnects.

Step 4: Verify downstream API refresh-token rotation

For custom internal integrations calling Calendly's API, remember that Calendly rotates refresh tokens on every exchange at POST /oauth/token. If two concurrent worker processes attempt to refresh using the same token without a distributed lock, the second worker triggers invalid_grant and permanently revokes the token family.

Detecting silent token failures automatically

Expired calendar tokens leave an unmistakable telemetry fingerprint: instantaneous cliff-edge collapse. Real client bookings consume slots one by one over days (18 slots → 15 → 12 → 9). An expired OAuth token drops 18 open slots straight to 0 between two consecutive 5-minute polls — often at 11:00 PM when an automated password-expiry policy triggers.

AvailTrack catches both OAuth layers automatically: if a host's Google or Microsoft token breaks and zeroes out their public Calendly slots, your zero-availability or next-slot rule emails the operations owner within minutes; and if the Calendly API token itself is revoked, AvailTrack flags the unhealthy connection directly. Read how AvailTrack handles OAuth token health and compare check frequencies on our pricing page.

Frequently asked questions

Does changing a Microsoft 365 or Google Workspace password disconnect Calendly?
Frequently, yes. Under strict enterprise identity policies — such as Microsoft Entra Continuous Access Evaluation or Google Workspace security resets that revoke active OAuth refresh tokens on password change — Calendly's background calendar token is invalidated immediately until the staff member signs in and re-authorizes Calendar Sync.
What happens to a public Calendly booking page when the host's calendar token expires?
When Calendly cannot verify free/busy conflicts on a host's primary connected calendar, it typically fails closed to prevent double-booking — causing the public event type to display zero available times across the month. If a secondary or delegated sub-calendar loses permission instead, sync can stall silently and allow double-bookings.
Do already-scheduled meetings get canceled if a Calendly OAuth token expires?
No. Existing appointments that were already written to Outlook or Google Calendar remain on both participants' calendars. However, new bookings stop working, invitee reschedule links fail to find open times, and automated calendar updates on cancellations stop propagating.
How can a firm administrator spot expired OAuth connections across 20 staff members?
Admins can inspect user sync statuses manually inside Calendly's Admin Center or watch for sudden cliff-edge drops in polled slot availability. When an advisor's 14-day slot count plunges from 16 slots to 0 inside a single 5-minute check cycle during non-booking hours, a broken calendar token is the prime suspect.

Catch broken calendar sync in minutes, not weeks.

When a password reset or M365 policy zeroes out a staff calendar, AvailTrack emails you immediately.