What changes when you move your workflows from Zapier to n8n?
Five things change: who keeps the workflows running, what you're billed for, where your app connections live, who hears about a failure, and the webhook address your forms send to. I'd rebuild each Zap in n8n, run both on the same inputs until the outputs match, then switch one trigger at a time and turn each Zap off.
Your Zaps work, and the bill grows with every action they complete. Moving to n8n changes what you pay for, and it also changes who looks after the workflows when something breaks. Which platform to pick has its own comparison page. This one covers the move: what changes, what to test, and how to switch without losing a lead.
I build on n8n, so weigh what I say about it with that in mind.
What changes for you
| What changes | On Zapier | On n8n |
|---|---|---|
| Who keeps it running | Zapier runs your Zaps and their webhook addresses, so there's no server on your side. | On n8n Cloud, n8n handles hosting, updates and scaling. Self-hosted, the server and its maintenance are yours. |
| What you pay for | Zapier bills in tasks, one each time a Zap completes an action. Checking for new data uses none. | n8n's paid plans bill in executions, one per run of the whole workflow, however many steps it has. Runs you start by hand while testing don't count. |
| Your app connections | Each app is connected inside your Zapier account. | Each app is connected again in n8n, which stores its own credentials. That's the moment to put every connection under an account you own. |
| Who hears about a failure | Error notifications go to your Zapier account email by default. | A failed run is logged in n8n. The email or Slack alert comes from an error workflow that someone sets for each workflow. |
| The address your forms post to | Zaps that start from a webhook use a Zapier address, on Professional, Team and Enterprise plans. | Each webhook workflow gets a new n8n address, live once the workflow is published. Every form or app that posted to Zapier gets pointed at it. |
The handoff
Here the handoff is the move itself. Each workflow goes through six steps, and each step has its own way of losing something.
| Step | Who or what does it | Where it breaks |
|---|---|---|
| List every Zap | You, from the Zap list and the task usage tab in Zap history. Team and Enterprise plans can also export your Zaps as a file. | A Zap that runs once a month is easy to miss in a short date range. A Zap that reads a Zapier Table or starts from a Zapier Form needs a new home for that table or form too. |
| Rebuild in n8n | Whoever will look after n8n, connecting each app again as they go | The docs I read on both sides describe no way to bring a Zap into n8n, so each one is rebuilt by hand. A missed filter step changes what the workflow acts on. |
| Run both on the same inputs | n8n, either watching the same trigger with its sends held, or fed real inputs taken from Zap history | With both tools live and both allowed to send, each new lead gets two emails and two CRM records. |
| Compare outputs | A person, matching each Zap run to the n8n run for the same record | Zapier only guarantees up to 60 days of run history, so older runs may be gone. Checking only the usual case misses the input that breaks things. |
| Switch the trigger | You, or whoever owns the form or app that sends the data | A sender nobody remembered keeps feeding the Zap, and the move looks done when it isn't. |
| Turn the Zap off | You, in Zapier, once the Zap's history has gone quiet | For up to several hours after a webhook Zap is off, Zapier keeps answering its address as if all is well. After that, anything still sending there reaches neither tool. |
Where it breaks
The costliest mistake happens during the side-by-side run. Zapier's own duplicate check only looks within one Zap, so nothing on either side stops n8n and the Zap from both replying to the same lead. While both run, the Zap stays the one that sends, and n8n's outward steps go to a test address or wait for a person.
The second is a sender nobody remembered, like a form on an old landing page that still posts to the Zapier address. Once the Zap is off, those leads stop arriving anywhere, and nobody on your side is told.
The third shows up after the switch. On Zapier, an error email reached you by default. In n8n, set an error workflow on every workflow before its trigger moves, so a failure reaches a person as well as the log.
What to test before you switch
- Run the same real inputs through both and compare each output field by field: the record written, the message sent, and who it went to.
- Include the awkward inputs from Zap history, like a blank field, a repeat submission and a run that failed in Zapier.
- Send the same input twice and confirm one email goes out and one record is written.
- Break a step on purpose, like a revoked app connection, and confirm a named person gets the alert.
- After repointing a trigger, watch the Zap's history until new runs stop, and only then turn it off.
The source-of-truth and dedupe rule
- Each trigger has one owner at a time. Until the switch it's the Zap, and after the switch it's the n8n workflow.
- Runs are matched on the record's own id, like the form submission or the deal, so one Zap run lines up with one n8n run.
- Anything that sends to a person checks that id first, so a repeat or a replay sends nothing the second time.
- Until the outputs match, the Zap's output is the answer key. After the switch, n8n's run log is the record of what happened.
What stays human
Deciding when the outputs match closely enough to switch is a person's call, and so is the order the workflows move in. Someone who knows the business can tell a harmless difference, like a date written another way, from one a client would notice. The owner of each form signs off on repointing it, because they know what else sends there.
After the move, someone has to own n8n: read a failed run, fix a renamed field, and keep the alerts going to a person who will act on them. Whoever builds it, the handover should leave your data, the automations in your own accounts, and handover docs with you.
What I'd pick, per situation
| If you | I'd pick |
|---|---|
| Have short Zaps, a trigger and an action or two, and nobody technical to look after n8n | Stay on Zapier. Short Zaps complete few actions, so the task bill grows slowly, and n8n would add a tool nobody on staff can fix. |
| Have long Zaps that run on every lead, and someone technical, in-house or hired, to look after them | Move to n8n Cloud one workflow at a time, starting with the one where a mistake costs least. |
| Need client data to stay on a server you control | Self-hosted n8n, once a named person has agreed to update it and back it up. |
| Rely on Zapier Tables or Zapier Forms inside your Zaps | Move the table or form first, or leave those Zaps on Zapier. Running both tools is fine while each trigger has one owner. |
| Have Zaps that email or message leads and clients | Move them last, once the side-by-side run shows one send per record and the error alert reaches a named person. |
| Haven't settled which platform suits your business | Read the n8n vs Zapier vs Make comparison first, since moving means rebuilding every workflow. |
Questions
Can I import my Zaps into n8n?
Not as far as I can tell. The import and export docs I read on both sides describe no way to bring a Zap into n8n, so I plan on rebuilding each one and testing it. On Team and Enterprise plans, Zapier's export of your Zaps makes a good checklist for the rebuild.
Will n8n cost less than Zapier?
It depends on how many steps your workflows have and how often they run. Count the steps and the monthly runs of your busiest Zap, then price it on both pricing pages, using tasks per completed action for Zapier and one execution per run for n8n.
How long should both run side by side?
Until n8n has handled every kind of input the Zap sees, including the rare ones. For a workflow that runs on every lead, that can be a short window. For a report that runs at month end, it means a full month.
Do I lose my Zap history when I leave Zapier?
Zapier only guarantees up to 60 days of run history, so export the runs you want before you cancel. They export as a CSV or JSON file, and they make good test inputs for the rebuild.
- Zapier pricing: a task is used when a Zap completes an action, checking for new data uses no task, Webhooks are a Professional plan feature, and Filters, Paths, Formatter, Tables and Forms are part of ZapierREAD
- Zapier help, Trigger Zap workflows from webhooks: Webhooks by Zapier is on Professional, Team and Enterprise plans, each Zap gets its own webhook URL, and after a Zap is turned off its URL keeps returning a success response for up to several hours before it returns an errorREAD
- Zapier help, How Zapier handles duplicate data in Zap workflows: deduplication only checks within the same Zap, so two Zaps on the same form both run, and an action creates a duplicate record when the app allows itREAD
- Zapier help, View and manage your Zap history: run details show the data each step received and sent, the task usage tab lists the Zaps that ran in a date range, and Zapier can only guarantee a maximum of 60 days of Zap run dataREAD
- Zapier help, Export your Zap history: Zap runs export as a CSV or JSON fileREAD
- Zapier help, Import and export Zap workflows: on Team and Enterprise plans, export your Zaps as a JSON file to back them up or share them with other accountsREAD
- Zapier help, Manage notifications when errors occur in Zap workflows: error notifications go to your Zapier account email address by defaultREAD
- n8n pricing: plans are based on workflow executions, and an execution is a single run of your entire workflow, however many steps are in itREAD
- n8n docs, Understand executions: only production executions count towards the quota, and manual runs while building or testing don't countREAD
- n8n docs, Choose how to use n8n: on n8n Cloud, n8n handles hosting, updates and scaling; self-hosted, you provide and manage the infrastructure, maintenance is your responsibility, and technical expertise is requiredREAD
- n8n docs, Create and edit credentials: credentials are stored in n8n to connect workflows to outside services, and n8n tests each one when you save itREAD
- n8n docs, Handle errors gracefully: failed executions are listed under Executions, and an error workflow, set for each workflow in its settings, runs when an execution fails and can send email or Slack alertsREAD
- n8n docs, Webhook node: separate test and production URLs, and the production URL registers when you publish the workflowREAD
- n8n docs, Export and import: n8n saves workflows in JSON format, and you can export them or import JSON files into your n8n libraryREAD
Related
- n8n vs Zapier vs Make: which should a service business run on?What each one bills for, who can change a workflow, what happens when a step fails, and where n8n, Zapier and Make each fall short for a small service firm.
- Automations that fail without telling anyoneAn automation can stop for weeks before anyone notices. The usual causes, what not knowing costs, and what an alert should say when a step fails.
- What a reliable automation looks likeFive properties any automation should have, checkable by an owner who doesn't build: logged runs, named alerts, one send per approval, safe reruns, an owner.
- Questions to ask anyone building your automationsQuestions to ask an automation builder before you hire them, with what a careful answer sounds like and the answers that should worry you.
- WebhookA webhook is a message one app sends another the moment something happens. What that means in plain English, how it fails, and how a system catches it.
- Send-onceSend-once means one approval sends one email, however many times the button gets pressed. Why duplicates happen and how a system keeps the rule.
- How do I stop website form duplicates in HubSpot?Where the handoff from your website form to HubSpot makes duplicates or loses leads, which field is the key, what stays with a person, and what I'd pick.