This question comes up in almost every first conversation I have with a client: which automation platform should we actually be using. Zapier is usually the name people already know, because it has been around longest and markets itself well. Make comes up second, usually from someone who has read a comparison article. n8n is the one almost nobody has heard of until I bring it up, and it is also the one I recommend more often than the other two combined.
Here is the actual reasoning, not the marketing version.
Where Zapier and Make fall down
Zapier’s biggest structural limit is one people do not find out about until they hit it: a single workflow, what Zapier calls a Zap, is capped at 100 steps. For a simple “new form entry creates a CRM record” automation that limit never matters. For anything with real branching logic, several conditional paths, retries, error handling, multiple destinations, it becomes a genuine ceiling far sooner than most people expect. I have had to split single logical workflows across multiple Zaps purely to work around that cap, which adds fragility rather than removing it.
Make is a step up in flexibility and its visual canvas is genuinely well designed for following a complex flow at a glance. Where it tends to lose out is depth of visibility once something breaks, and the ability to log the operational detail that actually proves a workflow is doing its job, not just that it ran.
Why n8n is what I actually build in
n8n solves the two problems above and adds a few advantages of its own.
- No hard step limit. A workflow can be as simple or as genuinely complicated as the job needs. That matters more than it sounds, because it means the automation gets built to match the actual process rather than the process getting simplified to fit the tool.
- Connects into a very wide range of software. Whatever is already in your stack, there is a strong chance n8n has a native node for it, and where it does not, it can usually reach it directly via API.
- No coding experience required to build with it. It is a visual, node-based builder. You do not need to be a developer to put a workflow together, though the ceiling is high enough that a developer can do genuinely advanced things in it too.
- Flexible triggers. A workflow can be started manually with a button press, by a form submission, by a webhook call from another system, or on a schedule. That flexibility means the same platform covers both “run this every night at 2am” and “run this the instant a customer submits a return request.”
- Proper visibility once it is live. You can log things like time saved per run, and get into the granular detail of exactly how a workflow executed, step by step, rather than just seeing a pass or fail. When something does go wrong, that detail is the difference between a five minute fix and an afternoon spent guessing.
That last point is the one clients undervalue until they have lived with a workflow for a few months. Being able to say, with evidence, how much time a specific automation has actually saved is what turns “we automated something” into a number you can put in front of whoever holds the budget.
Where I would not recommend n8n
To be fair to the other two: if you genuinely need one simple trigger-to-action automation, connect this form to that spreadsheet, and you never plan to touch it again, Zapier’s simplicity is a legitimate advantage and the setup time is shorter. The recommendation changes as soon as there is more than one step of real logic involved, which in practice is most of the automation work I end up doing for clients.
The honest summary
Zapier for the genuinely simple, one-off connection. Make if you specifically want the visual canvas and your workflows stay moderate in complexity. n8n for everything else, which in my experience is most of what actually moves the needle for a support or operations team. It is rarely the platform people expect me to say, but it is consistently the one that is still working well a year later rather than something someone has to quietly rebuild.
Choosing the right platform is one decision out of many when you are planning your first automation project, and getting it wrong early is expensive to unwind later. I cover exactly this kind of tooling decision, along with how to sequence a first 90 days properly, in The Automation Roadmap, my Maven workshop. Details and dates are here.