A website enquiry is saved in your form system, but no contact appears in the CRM and nobody receives the follow-up task. Zapier reports an expired app connection. Signing in again is the first repair; you still need to establish which enquiries stopped moving and what happened to them.
For a UK business using Zapier between its website, CRM and booking tools, treat this as two jobs: restore the connection, then reconcile the interruption. Keep somebody checking the original enquiry source while you investigate so new customers are not left waiting.
Confirm where the connection failed
Choose one missing enquiry and find its original submission time and reference. Open the relevant Zap’s history and look for that event, widening the date range and removing restrictive filters if necessary. Inspect the individual steps rather than relying on the Zap being switched on.
Use the evidence to choose your next move:
| What you find | What to check next |
|---|---|
| An authentication error, often 401 | Identify the app connection used by the failing step and test it. |
| A permission error, often 403 | Check whether the connected account can access the required record, workspace or action. |
| An On hold run | Read its stated reason before changing the connection. |
| No matching run | Check the trigger connection and original source records; there may be nothing in Zapier to replay. |
Zapier’s error guide distinguishes missing authentication from insufficient permissions. It also explains that repeated errors can switch a Zap off. Reconnecting will not resolve every permission or data problem.
An On hold status is not proof of expiry. Disconnected apps, task limits, plan changes and other restrictions can cause it. Zapier’s held-run guidance says it retains trigger data for held runs and directs users to reconnect a disconnected app before replaying them. That does not establish that every enquiry during the interruption reached Zapier.
Reconnect the correct account and check its reach
Start with the connection selected in the failing step. A connection to the right platform can still belong to the wrong business account or workspace.
For an ordinary personally owned connection, follow Zapier’s app-connection instructions:
- Open App Connections and identify the affected connection.
- Check its owner and the workflows associated with it.
- Use Reconnect and complete the sign-in with the intended app account. This option is available for connections you own.
- Use Test connection to check its status afterwards.
Ask the connection owner to complete their own sign-in if it belongs to somebody else. Before they do, confirm the business account required and who will verify the result.
Reconnecting an existing connection applies across the Zap workflows and other Zapier products that use it. Replacing it with a different account also changes the account used by those automations. Check that wider use before making a replacement: another workflow may depend on a different mailbox, workspace or set of records.
A successful connection test is useful evidence. Your acceptance test should still be a complete enquiry reaching the intended destination, with the right details and follow-up.
What if the person who connected it has left?
Establish who controls the app account and the Zapier connection before choosing a replacement. Record which workflows need to move and test each affected route afterwards. Avoid rebuilding several automations at once just to get past a login problem.
There is a separate option for Enterprise accounts: managed connections belong to the organisation and are created and shared by admins. If your business uses these, involve the responsible admin. The instructions for a personally owned connection are not the whole process.
Prove that a new enquiry can get through
Once access is restored, check that the intended Zap is on. Submit a clearly identifiable test through the real website route using contact details you control. Use a submission that meets the workflow’s normal conditions.
Trace that test through the original submission record, the Zap run and the destination. Check the specific outcome you rely on: a new enquiry associated with the correct contact, an assigned task, or a notification reaching the person responsible.
For example, if a quote form should create a contact and a sales task, finding the contact alone is incomplete verification. Confirm that the task exists, belongs to the right person and contains enough information to respond.
If no run appears, return to the trigger. Zapier’s trigger troubleshooting covers the connected account, permissions, selected event and trigger options. It also warns that turning a Zap on does not make existing data trigger automatically. Do not assume older enquiries will arrive simply because a fresh test succeeds.
Recover recorded runs using the right replay method
Before replaying anything, check the CRM and any sending or booking system for work that already happened. A colleague may have entered the enquiry manually, replied to the customer or arranged an appointment during the interruption.
Zapier offers different replay routes. Its explanation of replay makes an important distinction:
- Replaying errored steps resumes failed work without repeating earlier successful steps in the documented workflow.
- Replaying an entire Zap creates a new run and repeats every step, including steps that previously succeeded, using the current published workflow.
Choose the route after inspecting the original run and the destination. Replaying a whole workflow could send an acknowledgement or create a record again. A failed action may also have an uncertain outcome, so check the receiving system before repeating it.
Autoreplay does not replay On hold runs. Replay also has limits: the Zap must be on, sufficient tasks must be available, and the documented replay window is 60 days from the initial trigger event. Significant workflow changes can prevent failed-step replay. Filters and Paths are not rerun by that method. These limits do not promise that every historical record will be available or recoverable.
The current replay instructions allow manual replay of errored runs on Free; held runs require a paid plan. In Zap history, filter and select the relevant runs before using the replay control. Begin with one confirmed missing enquiry and inspect its result before processing a larger group.
If the available replay route does not fit your account or the run has changed too much, choose a controlled manual recovery rather than repeatedly clicking different replay options.
Find enquiries that have no Zap run
Use the original enquiry source to build the recovery list. Start just before the last confirmed successful transfer and finish after the first verified new submission following the repair.
Compare each source reference with the destination record. Record whether the enquiry arrived, whether somebody has already responded and what action remains. A matching contact email alone may be insufficient: an existing customer can submit a separate enquiry.
For anything genuinely missing, decide whether to enter it manually or use a supported import or transfer. Preserve the original enquiry time and reference where your systems allow it. Check whether the recovery method will also trigger messages, tasks or other workflows before using it in bulk.
Prioritise outstanding customer needs. Someone who requested a quotation several days ago needs a useful response from a person; recreating an old automated acknowledgement may add confusion. Assign each unresolved enquiry to somebody who will finish the follow-up.
Make the next interruption easier to spot
Give each important automation a named maintainer, a connection owner or responsible admin, and a simple description of its expected business outcome. Keep a short recovery note explaining where the original enquiries are stored and how to match them with destination records.
On Enterprise, Zapier’s Alerts page provides a central view of workflow issues and marks disconnected apps as Critical. Its availability is plan-specific, so do not build your monitoring process around a page your account cannot access.
Whatever the plan, agree who checks failures and how the business will notice missing outcomes. A regular comparison between source enquiries and assigned CRM work can reveal a gap that a connection-status check alone would miss. Repeat the real submission test after account handovers or changes to app access.
If you need help tracing the route from your website into the CRM, explore our AI and automation services or contact BuzzBoost. Tell us which apps are involved, when the last successful transfer happened and which step now fails. We can help define the repair, the recovery work and the checks needed to confirm enquiries are reaching the right person.
Featured image: AI-generated editorial artwork, not a photograph of a real BuzzBoost office, client or result.
