Contents
At a glance
- Category
- Automation mechanics
- Mechanism
- The source system sends an HTTP POST to a URL you publish
- Latency
- Effectively immediate, versus minutes for polling
- Delivery guarantee
- At-least-once - duplicates are normal, so idempotency is required
- Security
- Verify signatures where offered; otherwise a shared secret header plus payload validation
- Main risk
- Silent failure - nothing arrives and nothing errors. Add a heartbeat and a reconciliation sweep
What a webhook is
A webhook is an ordinary HTTP endpoint that you publish and hand to another system. When an event happens over there - a form is submitted, a payment succeeds, a call ends - that system sends an HTTP POST to your URL carrying the event details, usually as JSON. Your automation runs immediately, because it was told rather than because it asked.
Every automation platform gives you webhook URLs as a trigger type, and it is the single most useful trigger there is. The alternative is polling, which means checking on a schedule and accepting the delay and the wasted requests that come with it.
Setting one up
Four steps. Create the webhook trigger in your platform, which generates a URL. Register that URL in the sending system's settings, choosing which events it should send. Trigger a real event and inspect what actually arrives - never build against the documentation's example payload, because real payloads routinely contain extra fields, different nesting and null values the docs omit. Then build the rest of the workflow against that captured sample.
Most platforms distinguish a test URL from a production URL, and forgetting to swap them is the most common reason a workflow that worked yesterday receives nothing today. It is worth checking that first, before anything else.
Security, retries and delivery guarantees
Your webhook URL is a public endpoint that anyone who learns it can call. Three defences: verify the sender's signature where one is offered, which is the proper fix and is supported by most serious providers; check a shared secret in a header where it is not; and validate the payload shape before acting on it. Never trust the body's claim about which account it concerns without checking.
Delivery is at-least-once, not exactly-once. Senders retry when they do not get a fast 2xx response, so duplicates are normal and expected - which makes idempotency a requirement rather than a refinement. Respond quickly, ideally acknowledging receipt before doing the slow work, because a sender that times out will resend. And decide what happens when your endpoint is down: some providers retry with backoff for hours, others drop the event silently after one attempt, and the difference determines whether you need a reconciliation job.
What you should know
Webhooks fail quietly. Nothing arrives, no error appears anywhere, and you find out when a customer asks why nobody called back. Build a heartbeat: if no webhook has arrived in a period where one always does, alert. Pair it with a periodic reconciliation sweep that catches anything missed - a scheduled job that looks for records created in the source system with no matching record on your side.
Also log the raw incoming payload before parsing. When something goes wrong at 6am, the raw body is the only artefact that tells you whether the sender changed its format or your parsing broke. That is the same discipline as observability in an agent, applied to the boundary of your system.
Webhook FAQ
What is a webhook in simple terms?
A URL you give another system so it can tell you the instant something happens. Instead of your automation checking every five minutes whether a form was submitted, the form service sends a message to your URL the moment it is, and your workflow runs straight away.
How do I set up a webhook?
Create a webhook trigger in your automation platform to get a URL, register that URL in the sending system and pick the events you want, then fire a real event and inspect exactly what arrives. Build against the captured payload, not the documentation example - they differ more often than not.
Are webhooks secure?
The URL is public, so anyone who learns it can call it. Verify the provider's signature header if one exists, otherwise require a shared secret and validate the payload shape before acting. Never act on the payload's claim about which account it concerns without checking it independently.
Why am I getting duplicate webhooks?
Because delivery is at-least-once. Senders retry when they do not receive a prompt success response, and a slow endpoint gets retried while it is still working. Make the handler idempotent using the event ID so a repeat is recognised and ignored rather than processed twice.
What happens if my webhook endpoint is down?
It depends entirely on the sender. Some retry with exponential backoff for hours, some retry a few times, some drop the event after one failed attempt with no notification. Check the specific provider's policy, because it determines whether you need a reconciliation job to catch what was lost.
Why did my webhook stop working?
The usual suspects, in order: the workflow was left on the test URL rather than the production one, the workflow was deactivated, the sending system's registration expired or was removed, or the payload format changed. Logging the raw incoming body is what lets you tell these apart quickly.
Primary source
n8n - webhook node documentation
Checked August 2026. Anything priced or rate-limited moves, so verify against the source before you budget.
Want this built rather than explained?
Knowing what webhook means is one thing; having it running in your business is another. We scope one workflow at a time, quote a fixed price in writing after a free consult, build it, and hand it over with documentation you own.
