Opinion
A Zapier Alternative for When All You Need Is the Webhook
Most webhook automations are one step long. Something happens over there, and a message or an API call should happen over here. A workflow platform can do that, but you carry the whole platform in order to do it.
17 November 2026 · 4 min read
The job, stated plainly
Strip away the vocabulary and a large share of "automations" are the same four moves:
- Catch a webhook from a service.
- Decide whether this particular event matters.
- Reshape it into what the next system expects.
- Send it there.
A failed Stripe payment becomes a Slack message. A published CMS entry becomes a site rebuild. A form submission becomes a row in a table. One request in, one request out.
What comes along with a workflow platform
Zapier, Make and n8n are built for much more than that, and they are good at it: hundreds of app connectors, multi-step flows, branching, schedules, lookups, stored state. The price of that range is everything you take on before your first event moves:
- An account and a workspace, with members, roles and billing to look after.
- App connections. You authorise each service, and those authorisations expire, break and belong to whoever set them up.
- A step editor. Even a one-hop flow is a trigger step, an action step and a field-mapping screen for each.
- A meter. These platforms typically charge by the task, operation or execution, so a noisy webhook is a cost you have to watch.
- A place where logic hides. Six months later, somebody has to remember that the reason sales gets a ping is a flow in somebody's personal workspace.
None of this is a flaw. It is weight, and it is worth carrying when you use what it buys. For a single hop, it is mostly overhead.
The light version
On webhook.stream, the same job is a rule attached to a URL. You get the URL with one click, without an account. The rule has three parts.
When
| Field | Operator | Value |
|---|---|---|
body.type | equals | invoice.payment_failed |
body.data.object.amount_due | is greater than | 5000 |
Turn it into
{
"text": ":warning: Payment failed for *{{body.data.object.customer_email}}*\n<{{body.data.object.hosted_invoice_url}}|Open invoice>"
}Send it to
A Slack incoming webhook, a Discord channel, a Telegram bot, or any HTTPS URL with the headers it needs.
You write the template against a real captured request, and the editor shows exactly what would be sent as you type. Every delivery is logged next to the request that caused it, with the status code, the timing and the start of the reply. That is the whole product surface for this job. There is no second screen.
Side by side
| Workflow platform | webhook.stream automation | |
|---|---|---|
| Before the first event | Account, workspace, app connections | One click for a URL |
| Trigger | Webhooks, polling, schedules, app events | An incoming HTTP request |
| Shape of a flow | Many steps, branches, loops | One request out per automation; several automations can run on the same request |
| Reshaping data | Field mapping, formatters, code steps | A template with placeholders and a fixed set of filters |
| Destinations | App connectors and HTTP | Any public HTTPS URL, with your headers |
| Failed deliveries | Usually retried or replayable | Logged, not retried; pauses after 15 in a row |
| Cost model | Usually per task or operation | Free to build; $3 or $19 once to keep a URL |
Where the light version stops
It is better to know this before you build than after:
- One hop. An automation sends one request. It cannot take the reply and feed it into a second step.
- No lookups. It cannot fetch a customer record to enrich a message. It works with what the webhook carries.
- Templates, not code. No loops, no arithmetic, no if/else inside a template. Two automations with opposite conditions are your if/else.
- Webhooks only. Nothing runs on a schedule and nothing polls.
- Static credentials. The destination has to accept a URL or a header you can paste. There is no OAuth sign-in flow.
- At-most-once delivery. A failed delivery is recorded and shown, not retried.
How to choose
Use a relay when the flow is "webhook in, request out", the destination speaks HTTPS, and a missed notification would be an annoyance and not an incident.
Use a workflow platform when you need several steps in sequence, data from a second system, a schedule, an app that only offers OAuth, or guaranteed retries.
Plenty of teams end up with both: a platform for the three flows that are real workflows, and a relay for the thirty that were only ever a webhook that needed to arrive somewhere else in a different shape.
Five single-hop jobs, written up
- Stripe payment failures and disputes into Slack
- GitHub releases and CI failures into Discord
- Shopify orders to Telegram or a phone push
- Any webhook into a GitHub Actions run
- Any webhook into an Airtable row
Try the single hop
Create a URL, send it one real event, and build the rule against that event. If it turns out you needed a workflow after all, you have lost ten minutes and gained a captured sample payload.
Questions
Is webhook.stream a full replacement for Zapier, Make or n8n?
No. It replaces one thing those tools do: reacting to an incoming webhook by sending an HTTP request somewhere else, optionally filtered and reshaped. It has no app directory, schedules, multi-step workflows or lookups. If your automation is a single hop, it is quicker to set up and easier to reason about.
Do I need an account?
No. A webhook URL is created with one click and automations are configured on that URL. There is nothing to sign up for and no workspace to manage.
How is it priced?
Building and testing is free. A free URL lives for up to 7 days and can make 1,000 automation deliveries a day. To keep a URL for longer, a one-time pass costs $3 for 30 days or $19 for 1 year. Nothing is metered per event.
What happens if the destination is down?
The attempt is logged with its status and timing, and it is not retried. After 15 failures in a row the automation pauses and tells you why. For work that must never be lost, deliver to a queue or to your own service.
Keep reading
The webhook glue server you keep rewriting
Every team has a tiny service that receives a webhook, reformats it and calls another API. The code is twenty lines; running it is the real cost. When to keep writing it and when not to.
ReadTransform a webhook payload
Reshape webhook JSON into the payload another API expects using a template: pick fields, keep types, set defaults and format values, with a live preview against a real request.
Read