Switching From Zapier to n8n: Migration Steps and Gotchas

Switch from Zapier to n8n with this step-by-step migration guide: inventory, rebuild, webhook and credential gotchas, parallel runs, and a clean cutover.

By Galactic News Editorial · Published 2026-11-08T16:23:21+00:00 · Updated 2026-11-08T16:23:21+00:00

To switch from Zapier to n8n, you rebuild each Zap as an n8n workflow by hand. There is no import button. n8n cannot read Zapier exports. The move is still worth it for many users, because the billing model is different and the software can run on your own server. This guide covers the full migration: inventory, setup, rebuild, testing, and cutover.

The core trade is simple. Zapier charges per task, so a Zap with five actions costs five tasks per run. n8n charges per execution, so a workflow with fifty steps costs one execution per run. Self-hosted n8n costs nothing in licence fees. You pay only for the server and your time to maintain it.

Zapier vs n8n at a glance

Zapier has far more native integrations. n8n closes the gap with its HTTP Request node, which can call any API. The billing difference is where the money is. A five-step Zap that runs 1,000 times a month uses 5,000 Zapier tasks. The same work in n8n uses 1,000 executions.

Step 1: Inventory your Zaps before you switch from Zapier to n8n

Do not rebuild blind. List every active Zap first. For each one, record the trigger, the apps, the step count, and how often it runs. Note whether it touches money, customers, or regulated data.

Zapier lets you export your Zap data from account settings. Go to Settings, then Security and data, then Download my Zap workflows. You get one file with all your Zaps. It is useful for building the inventory, but n8n cannot import it.

Pull your task usage report too. Rank Zaps by tasks used. The biggest consumers are your best migration candidates. They save the most money when they move.

Sort the list into three groups. Migrate the simple, high-cost Zaps first. Migrate the complex ones later. Leave Zaps that rely on a Zapier-only app or feature until last, or keep them on Zapier.

Step 2: Choose Cloud or self-hosted n8n

n8n Cloud is the fast start. Sign up for a free trial, with no credit card required, and build in the browser. Starter costs 20 euros per month billed annually and includes 2,500 executions per month. Pro costs 50 euros per month billed annually. Business costs 667 euros per month billed annually and is self-hosted only.

Self-hosting runs the free Community Edition on your own server. The software is free under the Sustainable Use License. You pay for the server, backups, and your time to patch it. n8n recommends self-hosting for expert users. Its docs warn that mistakes can lead to data loss, security issues, and downtime.

Pick Cloud if you want n8n to handle hosting, updates, and scaling. Pick self-hosting if you want full control of your data and no per-execution bill. Either path works when you switch from Zapier to n8n; the choice is about who maintains the server.

Step 3: Rebuild your first Zap as a workflow

Rebuild, do not transliterate. A Zap and an n8n workflow share concepts, but the structure is different. Map the concepts first, then rebuild cleanly.

1. Map the concepts. A Zap becomes a workflow. A trigger becomes a trigger node. Each action step becomes a node. A Formatter step usually becomes an expression or a Set node. A Code step becomes a Code node.
2. Pick the trigger node. Use a Webhook node for incoming data, a Schedule Trigger for timed runs, or the app trigger node for the service in question. n8n workflows can have multiple triggers on one canvas.
3. Connect the nodes. Drag nodes onto the canvas and link them in order. Three chained Zaps that pass data through Storage often become one n8n workflow.
4. Re-enter credentials. OAuth connections cannot transfer. Every app must be connected again in n8n. Find the account owner for each one before you start.
5. Replace webhook URLs. Every Catch Hook URL on hooks.zapier.com must be replaced with an n8n Webhook node production URL in the sending system. The n8n test URL only works while you are listening in the editor, so never paste it into a live system.
6. Test with real data. n8n lets you run each step on its own. Execute the workflow with a real record and check every output.
7. Activate and watch. Turn the workflow on and monitor its first runs in the execution log. Fix errors before you move to the next Zap.

A simple three-step Zap often takes under an hour to rebuild. A complex workflow with branching and formatting can take several hours. Most teams finish a full migration in two to four weeks.

Gotchas when you switch from Zapier to n8n

Five traps catch almost every migration. Know them before you start.

Data arrives as arrays, not flat text

Zapier flattens line items into comma-separated strings. n8n keeps them as structured JSON arrays. Every downstream node runs once for each item in the array. A workflow that worked in Zapier can send five emails in n8n where it sent one before. Check each node output for item counts.

Docker defaults to UTC

Zapier handles timezones from your account settings. A self-hosted n8n container defaults to UTC. A schedule trigger set for 8:00 in your local time will fire at 8:00 UTC instead. Set the timezone in your container configuration before you activate scheduled workflows.

Your history does not move

Run history stays in Zapier. If you need it for audits, export it before you cancel the plan. Start early on accounts with a long history.

Polling and webhooks behave differently

Some n8n trigger nodes poll on a schedule. Others wait for a webhook. Check which kind each trigger uses. A polling trigger on a five-minute interval will not react instantly, even if the Zapier version did.

Premium apps need a native node or an API call

Zapier lists more than 9,000 apps. n8n lists more than 400 integrations. Most major apps have native n8n nodes. For the rest, use the HTTP Request node to call the app API directly. That takes more setup, so budget extra time for unusual apps.

Step 4: Run both systems in parallel

Never cut over cold. Run Zapier and n8n side by side for one to two weeks. Route the same events to both systems and compare the outputs.

For webhooks, duplicate the payload to both endpoints. A reverse proxy or an edge worker can do this. For scheduled workflows, let both run and check that the results match. This parallel run is the safety net for anyone who wants to switch from Zapier to n8n without downtime.

Watch for duplicates. Two live systems can send two emails or create two records. Use test data and test accounts during the parallel run. Turn off the Zapier version only when the n8n version has run clean for several days.

Step 5: Cancel Zapier and clean up

Once every workflow is stable in n8n, cancel the Zapier subscription. Do it in this order. First, export any remaining run history. Second, turn off each Zap. Third, cancel the plan. Fourth, revoke Zapier access from connected apps if you no longer need it.

Keep the export file. It documents how your old automations worked. That record is useful when someone asks why a workflow behaves a certain way six months later.

Community tools that can help

There is no official importer, but community tools exist. One open-source tool, migrate-to-n8n, converts a Zapier export into an n8n workflow file with a report on each step. It runs locally on your machine. It converts simple steps well and flags the rest for manual work.

Treat any converter output as a draft. Import the file, pick your credentials in the nodes the report lists, and test before you activate. Converters cannot move credentials or history. They save typing, not thinking.

Bottom line

To switch from Zapier to n8n, inventory your Zaps, pick Cloud or self-hosting, and rebuild each workflow by hand. Watch the five gotchas: arrays instead of flat text, UTC timezones, lost history, webhook URLs, and re-authorized credentials. Run both systems in parallel for one to two weeks. Then cancel Zapier. The payoff is execution-based pricing and full control of your data. The price is your time, and for heavy users it is time well spent.