n8n or Make? Which automation fits your business?
n8n and Make both connect your tools, but with different strengths. What decides it is how complex your process is, how many systems are involved, who looks after it and what it costs – not the feature list.
When businesses want to automate processes between their tools, they quickly end up with two names: n8n and Make (formerly Integromat). Both connect apps like Gmail, Google Sheets, a CRM or Slack through visual workflows. Both can include AI steps.
So the question isn't which tool is "better" – it's which one suits you better.
The short version
| n8n | Make | |
|---|---|---|
| Hosting | cloud or your own server | cloud only |
| Billing (cloud) | per workflow execution | usage-based – every step counts |
| Getting started | somewhat more technical | very beginner-friendly |
| Custom code | JavaScript and Python directly in the workflow | possible, but more limited |
| AI agents | dedicated building blocks for AI agents | AI modules and agents available |
When n8n fits
- Data should stay in-house. n8n can run on your own server. For many businesses with sensitive customer or HR data, that's an important argument.
- Workflows have many steps. In n8n's cloud version, one execution counts once – no matter how many steps the workflow has. For large workflows processing many records, that can be considerably cheaper.
- You need logic beyond clicking. Code nodes let you handle edge cases cleanly instead of building them in roundabout ways.
When Make fits
- Your team should build small automations themselves. Make is very approachable: the interface is clear and many ready-made modules take work off your hands.
- The processes are manageable. For simple connections with few steps, Make is quick to set up.
- You don't want your own server. Make runs entirely in the cloud, with an EU data centre if you need one.
Four questions that decide it
- How complex is your process? A few steps, one clear path: Make is often perfectly enough. Many branches, edge cases and large volumes of data: n8n gives you more room.
- How many systems are involved? The more programs are involved, the more error handling and overview matter. Also check up front that the tools you use are supported on both platforms.
- Who looks after it afterwards? With n8n on your own server, running it comes on top: updates, backups, monitoring. Make takes that off your hands. In exchange you have less control.
- What does it cost to run? They bill differently (see the table). With few steps the difference is small, with large workflows and many records it can become noticeable. So calculate with your real process, not with the price list.
What else matters
The tool question usually isn't the decisive one. More important:
- Is the process clear? Automate an unclear process and you get an unclear process that makes mistakes faster. That's why I map processes first – for example with BPMN.
- Who looks after it afterwards? An automation needs someone who receives error alerts and can make small adjustments.
- What happens when something fails? Good workflows speak up when something goes wrong – instead of silently losing data.
My recommendation
For most projects with customer data, multiple systems and AI steps, I go with n8n – for the control over data and the flexibility. But if a team mainly wants to build simple connections themselves, Make is often the easier start.
Not sure what fits your process? In a free 15-minute call we'll look together at which approach suits your business. After that it continues in the automation pilot: process first, then the tool.
What such a project costs is covered in What does custom automation cost a small business? For examples of workflows, see 7 processes every marketing agency should automate.
