Connect webhooks to Zapier

This guide builds a Zap that adds new enquiries to Google Sheets. Arbour sends an event, Zapier fetches the enquiry, then checks for an existing row before adding one. Start in Settings → Integrations in Arbour.

Prerequisites

• An Arbour role with Manage outbound webhooks and Manage API keys.
• A paid Zapier account. Both Webhooks by Zapier and API by Zapier require a paid plan. API by Zapier is currently in beta, so its settings may change. See Zapier's API connection guide.
• A Google account with editing access to a test spreadsheet. Put Enquiry ID, Name, Stage, Event date and Created at in cells A1 to E1. Use fictional enquiry details while testing.
• An Arbour API key, saved in a Zapier connection as described below.

Connect Arbour

1. 01

   Create a Zap. Select Webhooks by Zapier as the trigger and Catch Hook as the event. Leave Pick Off A Child Key blank so the outer event type remains available.

2. 02
   Open the trigger's Test tab and copy the webhook URL.
3. 03

   In Arbour, press Add endpoint under Webhooks. Name it Zapier enquiries to Sheets, paste the URL and subscribe only to pipeline_item.created. Create the endpoint and save its signing secret in your secure credential store. It is separate from the API key.

Zapier parses the JSON into fields for later steps. Its webhook trigger guide describes the setup.

Test receipt

1. 01

   Open the endpoint in Arbour and press Send test ping. A successful response checks connectivity. The ping has empty data, no enquiry fields and no entry in Arbour's delivery log.

2. 02

   Create an enquiry with fictional details and an event date. Return to Zapier and press Test trigger. Select the sample whose outer type is pipeline_item.created and whose data.type is enquiry.

3. 03

   Check that data.pipelineItemId contains your enquiry's ID. If only the ping appears, create another test enquiry and test the trigger again before mapping fields.

The payload examples show a copyable ping and enquiry event. Their IDs are fictional; use the live sample for the API request. The webhook carries identifiers, not the enquiry's name or event date.

Build the automation

Keep only new enquiries

Add Filter by Zapier immediately after the trigger. Set two conditions joined with AND, using (Text) Exactly matches:

Zapier may display nested fields as Data Type and Data Pipeline Item Id. Check the source path against your sample. The filter rules reference explains the matching options. Both conditions must pass, so pings and newly created bookings stop here.

Fetch the enquiry

1. 01

   In Zapier's Apps page, choose Add connection → API by Zapier. Select Static Headers (API key). In Headers, enter Authorization: Bearer arb_..., replacing arb_... with your complete Arbour API key. Name the connection Arbour and set Domain filter to api.usearbour.com. Leave the optional Test URL blank and save.

2. 02

   Back in the Zap, add API by Zapier → API Request after the filter. Choose the saved Arbour connection and set Method to GET. Enter https://api.usearbour.com/v1/pipeline-items/ in URL, then insert the trigger's data.pipelineItemId field after the final slash.

3. 03

   Leave query parameters, extra headers and body empty. Press Test step and inspect the JSON response. It should contain your enquiry's id, displayName, stage.name, startDate and createdAt. Select these fields from this API response in later steps.

The connection supplies the header automatically. Do not add X-Arbour-Org; the key belongs to one organisation. Keep the API host fixed and map only the record ID from the webhook. See Zapier's API Request instructions for the request fields.

The response is the enquiry's current state. Its name, stage or date may have changed since the event was sent. startDate can be null.

Check for a row, then add one

1. 01

   Add Google Sheets → Lookup Spreadsheet Row. Connect your Google account, then choose the test spreadsheet and worksheet. Set Lookup column to Enquiry ID and Lookup value to id from the API response. Leave supporting lookup fields empty.

2. 02

   Set Successful if no search results are found? to True. Keep automatic row creation off. If multiple results are found, return the first result. Test with the sheet still empty.

3. 03

   Add another Filter by Zapier. Continue only when the lookup's Zap Search Was Found Status is (Boolean) Is false. A successful search step does not by itself mean a row was found. Confirm that this filter passes for an absent Enquiry ID.

4. 04

   Add Google Sheets → Create Spreadsheet Row and select the same spreadsheet and worksheet. Map each column using the earlier API Request response, not the lookup's empty output. Test the action once and check the new row.

These are Zapier's Google Sheets actions. Its search-action guide explains the no-result setting and found-status field.

Use mapped fields, not the literal field names or values copied from the test. Keep Event date optional; do not replace a missing date with today's date or the text null.

In the row-creation action, type a single apostrophe ' before the mapped Name and Stage fields. This makes Sheets treat those values as text, including names that start with a formula character. See Zapier's guidance on text being interpreted as formulas.

Turn it on

1. 01

   Publish the Zap and confirm it is on. Check that the Arbour endpoint is enabled and still contains this Zap's URL. Zapier uses the same receiver URL for testing and live delivery.

2. 02

   Create a second fictional enquiry. Open its run in Zap History and follow the trigger, filter, API request, row lookup and row creation. Check all five cells in the sheet.

3. 03

   Re-test the lookup using that same enquiry ID after the row exists. It must find the row, and the following filter must stop. Do not test the row-creation action on its own again, because doing so bypasses the duplicate check.

4. 04

   Send another ping and create a fictional booking. Neither should reach the API request or add a row. Also test an enquiry without an event date and confirm its date cell is blank.

A lookup before insertion reduces duplicates when a fresh run searches after an earlier run has finished. Two simultaneous runs can both see no row and add one. This is not an exactly-once guarantee.

Zapier replays reuse search results rather than repeating the lookup. If a row write may already have succeeded, check the sheet before replaying a run. A replay with an old empty lookup result can add a duplicate. See the limitations in Zapier's search-action guide.

Troubleshoot

Arbour's delivery status records Zapier's acknowledgement. A delivered webhook does not prove the API request or spreadsheet action finished. Inspect Zap History for downstream failures. Arbour does not automatically retry a request that Zapier accepted; Arbour's delivery guide explains its retries and manual redelivery limits.

Zapier documents receiver URL ownership and delayed 404 responses in its webhook trigger guide.

Advanced security

Catch Hook parses the body. For signature verification, Catch Raw Hook exposes the unparsed body and request headers, with a 2 MB limit. Changing triggers requires retesting and remapping fields; raw capture alone does not verify anything. See Zapier's raw-hook option.

Follow the shared signing reference: use the complete signing secret, the timestamp header and exact raw body for HMAC-SHA256; compare the v1=<hex> signature, reject stale timestamps and deduplicate delivery IDs durably. Do not parse and rebuild JSON before verification.

A verification step inside the Zap runs after Zapier has acknowledged receipt. To reject invalid requests before acceptance, put a verifying HTTPS receiver in front of the private Zap URL. It must verify first, durably queue accepted work and respond within Arbour's ten-second deadline. Forward only accepted events to Zapier. Building and hosting that receiver is separate from this guide.