n8n vs Make: which one should you choose to automate your operation?
The choice is not about price or available nodes. It is about who will maintain it 18 months from now.
n8n and Make (formerly Integromat) look the same at first glance: both let you connect APIs without writing code, both have hundreds of integrations, and both charge by execution. The operational reality is very different, and the right decision depends on where you want to be in 18 months, not on what looks nicer today.
The real differences (not the marketing ones)
Make is pure SaaS: you pay, they host, and you build workflows in their visual interface. n8n is open source and self-hostable: you can run it on your own server for US$15/month instead of paying US$300+ to Make when volume grows.
That difference feels minor when you start (Make does have better UX). It becomes critical when you reach 50,000+ executions per month, when your data needs to flow through servers in your legal jurisdiction, or when your end client requires that there is no third-party SaaS in the middle.
| Dimension | n8n | Make |
|---|---|---|
| Hosting | Self-hosted or own cloud | SaaS only |
| Entry price | US$0 (self-hosted) / US$20 cloud | US$10 / month |
| Price at 100k executions | ~US$60-120 (hosting) | US$340+ |
| Native integrations | ~400 | ~1,500 |
| Code nodes (JS/Python) | Native, no limit | Limited on free plan |
| Learning curve | Medium-high | Low |
| Error handling | Deep, with customizable retries | Good but less flexible |
| Compliance / on-prem | Yes (self-hosted) | No |
When Make is the right answer
- A marketing or operations team without a technical profile will maintain the workflows.
- Total volume is low (< 20,000 executions/month) for the foreseeable future.
- You need integrations with rare vertical apps (Make has 3-4x more native connectors).
- You want ZERO infrastructure: no container, no VPS, nothing.
When n8n is the right answer
- Someone technical (in-house or agency) can maintain a container or VPS.
- You expect more than 50,000 executions/month within 6-12 months.
- You need complex code nodes (non-trivial data transformations, internal API calls, deep conditional logic).
- Compliance/legal requires the data to remain on your servers or a specific jurisdiction.
- You want to version workflows in Git (n8n exports JSON; Make does not do this natively).
Real 12-month cost (typical scenario)
Logistics company with 7 automated processes, ~80,000 executions/month:
- Make: Teams plan + extra operation packs = ~US$420/month = US$5,040/year. Add operator hours to maintain it: ~10h/month at US$25/h = US$3,000/year. Total: US$8,040/year.
- n8n self-hosted: 4 vCPU VPS on Hetzner = US$25/month = US$300/year. Maintenance hours (higher than Make): ~15h/month at US$25/h = US$4,500/year. Initial agency setup: US$2,500 once. Year 1 total: US$7,300. Year 2+: US$4,800/year.
With n8n you save ~40% starting in year 2, but you need the technical person who maintains the server. If you do not have that person and do not want to hire them, the savings are fictional.
The common trap when choosing
The mistake we see most often: non-technical teams choose Make because the curve is easy, reach 60,000 executions in 4 months, the bill jumps to US$280/month, they try to migrate to n8n without a technical team, and they end up with 7 broken automations for 3 weeks while a freelancer rewrites them.
The right question at the start is not "which one works for me today"; it is "how much volume will I have in 12 months and who will maintain this?".
Can you use both?
Yes, and sometimes it makes sense. Make for simple workflows maintained by marketing (forms -> CRM, Slack -> Notion). n8n for critical operations or internal data automations (ERP -> reports, AI lead scoring, cross-system sync). Total cost does not double if the split is clear.
What about Zapier in this comparison?
Zapier is ~30-50% more expensive than Make at medium volumes, with simple UX and less technical power than either option. If Zapier already works and you do not have price or feature bottlenecks, do not migrate. If you are starting from zero, do not choose it over Make or n8n.
n8n cloud (paid version) vs self-hosted?
n8n cloud costs US$20-50/month for small volumes. It loses the cost advantage after 30k executions; from there, self-hosting always wins. But it saves maintenance. It makes sense for small teams without DevOps.
What we recommend
If your operations team does not have a technical profile and you will stay under 30,000 executions/month: Make, no doubt. You will lose more in n8n learning time than you save on the bill.
If your team is technical (or you have an agency maintaining infra), if you will exceed 50,000 executions/month in 6 months, or if you need workflow versioning: self-hosted n8n. The accumulated savings over 24 months usually justify the initial setup.