n8n vs Make vs Zapier: which should a small business actually use?

Use Zapier when getting started fast matters more than the bill, and your workflows are genuinely simple. Use Make when you need real branching and decisions, at a lower price per step. Use n8n when you need to keep data on your own server, or your volume has grown enough that per-step pricing has become your biggest line item. Most small businesses start on Zapier for good reasons, and only move when one of those reasons stops applying.

Here is the longer version, including where each one genuinely falls down.

What actually separates them

All three connect apps and move data between them. All three marketing sites will tell you they do everything. The differences that matter come down to three things: how they charge you, how much logic you can build before it gets painful, and where your data ends up.

Everything else — how many connectors they list, template galleries, AI feature announcements — is close to noise for a small business.

Zapier

The pitch: the biggest library of app connections, and the shortest path from idea to something that works.

What it is genuinely good at. If your workflow is "when this happens in app A, do that in app B", Zapier will have you running in fifteen minutes, and it probably already connects to the niche tool your industry uses. That counts for more than people admit. A workflow running today beats a better one you never finish.

Where it falls down. Two places, both predictable. First, the bill: Zapier charges per task, and a task is roughly every action in every run. A five-step workflow running 500 times a month is 2,500 tasks, and that climbs faster than most people expect. Second, logic: several branches, looping over a list, or reshaping data in the middle are all possible but awkward, and workflows get hard to follow once they grow past a certain size.

Pick it when you have a few simple workflows, modest volume, nobody technical in-house, and your time is worth more than the subscription.

Make

The pitch: a visual canvas with genuinely capable logic, at a lower price per unit of work.

What it is genuinely good at. The builder shows the whole flow as a picture, which makes complicated branching far easier to follow than a list of steps. Working through a list item by item and then combining the results is built in, rather than a workaround. Per-operation pricing usually works out cheaper than Zapier at the same volume, especially for workflows with a lot of steps.

Where it falls down. The learning curve is real. The same flexibility that makes it powerful makes it easy to build something confusing. The connector library is large but not Zapier's, so now and then you will be wiring up a request by hand. And error handling works well but has to be set up on purpose — the defaults will not save you.

Pick it when your workflows involve real decisions, your volume makes Zapier's per-task bill uncomfortable, and somebody on your side is willing to learn the tool.

n8n

The pitch: you can run it on your own server, and it charges per workflow run rather than per step.

What it is genuinely good at. Three things nothing else here offers together. You can host it yourself, so your data never passes through another company's automation service at all — decisive if you have rules about where data may live, or you handle anything sensitive. It charges per workflow run rather than per step, so a twenty-step workflow costs the same as a two-step one, which changes the maths on complicated jobs completely. And you can drop into code at any point, so there is no ceiling where you have to abandon the tool and start again.

This is the one I reach for most often, and the per-run pricing plus the self-hosting option is why.

Where it falls down. Hosting it yourself is a real job — updates, backups, keeping it online, and somebody to notice when the server falls over. Their hosted version removes that, but it also removes the reason about keeping data on your own infrastructure. The connector library is smaller than Zapier's, though that matters less than it sounds, because building a connection by hand is straightforward. And it is the most technical of the three: comfortable for a developer, harder going for someone non-technical.

Pick it when you need to control where data lives, your workflows are complex enough that per-step pricing hurts, or you want the option to extend with code without switching tools later.

A comparison you can act on

| | Zapier | Make | n8n | | --- | --- | --- | --- | | Charges you by | Task (each action) | Operation | Workflow run | | Cost when there are many steps | Rises steeply | Rises moderately | Stays flat | | Complex branching | Awkward | Strong | Strong | | Run it on your own server | No | No | Yes | | Custom code | Limited | Some | Full | | Time to first working workflow | Fastest | Middle | Slowest | | Good for a non-technical user | Best | Workable | Hardest |

The question that actually decides it

Before you compare a single feature, answer this: is there anywhere your data is not allowed to go?

If you handle health records, financial data, or personal data belonging to people in the EU or UK under a contract that limits who may touch it, that question narrows the field before anything else does. Self-hosted n8n keeps the data on machines you control. Zapier and Make both mean your data passes through their systems, which is completely fine for most businesses — but it should be a decision you made on purpose and wrote down, not something you discover during a client's security review.

If the answer is no, and for most small businesses it is, decide on cost and complexity instead. Start with whichever gets something working soonest.

What I would tell you on a call

Start with the simplest tool that can do the job. A working automation on a slightly-too-expensive platform beats a perfect one you never finished. Moving between these tools later is annoying but not a disaster, and you will understand what you actually need far better after six months of running something than you do today.

There are two cases where I would skip ahead to the more capable option. If you already know your volume will be high — thousands of runs a month — the per-task maths will bite soon enough that starting on Zapier is a false economy. And if you have a hard requirement about where data lives, build it right the first time rather than migrating into it.

The platform is also, honestly, one of the smaller decisions. Whether an automation works out depends on how well the underlying process was understood, how the odd cases are handled, and whether the people who have to live with it had a say in designing it. A well-chosen platform will happily host an automation nobody uses.

If you want a view on which of these fits your situation, the audit covers exactly that — including the case where the tools you already pay for do the job and you should not buy anything at all.