That’s the gap n8n is built to close.
n8n (people say “n-eight-n”) is an open source workflow automation tool. You build automations on a visual canvas by connecting nodes, most of the time without writing any code, and you can drop into JavaScript for the cases a pre-built node doesn’t cover. The basic idea is simple: something happens in one app, n8n notices, and it goes and does something in another app.
What makes n8n different from most of the automation tools you’ve probably heard of is that you can run it yourself. Instead of renting automation from a cloud vendor and paying per task, you host it on your own server and the cost stays flat no matter how much you lean on it.
A handful of terms show up constantly once you start using n8n, and it’s worth getting comfortable with them before anything else in this series.
A node is a single step in a workflow. It might call an API, send an email, or reshape some data. A trigger is just a special kind of node: the one that kicks a workflow off, whether that’s a schedule, an incoming webhook, or an event from a connected app. String a bunch of nodes together starting from a trigger and you get a workflow, which is the automation itself. Every time that workflow runs, that single run is called an execution, and n8n keeps a log of exactly what data moved through each node and whether anything failed.
Keep these four words in your head. Every post in this series leans on them.
People ask us this a lot, so here’s the actual answer instead of a sales pitch.
Zapier is genuinely the easiest place to start if you’ve never automated anything before. The interface holds your hand well. The catch is that the pricing model punishes you as your usage grows, and once your logic gets more than a couple of steps deep it starts feeling limited.
Make (it used to be called Integromat) has a nicer visual canvas for complex branching logic and handles multi-step automations more gracefully than Zapier does. It’s still a cloud-only, billed-per-operation product though, so the cost dynamics are similar.
n8n asks a little more of you up front. The interface isn’t quite as beginner-friendly on day one. What you get in exchange is a tool you can self-host, so your cost is the server you’re already paying for rather than a bill that climbs with every task you run, plus the ability to write real code inside a workflow the moment a built-in integration falls short.
If you’re automating something small and occasional, Zapier will serve you fine and there’s no reason to overthink it. If automation is going to be load-bearing infrastructure for how your business runs, which is where most companies we work with eventually end up, n8n tends to be the better long-term bet.
Here’s a workflow we build in some form for almost every client: someone fills out a contact form on a website, and instead of that submission sitting in an inbox until someone checks it, it goes straight to the team.
A webhook trigger catches the form submission the instant it happens, with all the field data attached. A Set node cleans that data up, maybe adds a timestamp or reformats a phone number. A Slack node then posts it straight into a #leads channel.
That’s it. Three nodes, and nobody has to remember to check anything. Most useful automations follow this same shape once you get used to seeing it: something happens, the data gets tidied up, and an action fires somewhere else.
Not every annoying task deserves a workflow, and building one for the wrong thing just adds maintenance overhead. Before we build anything for a client, we’re checking for three things: does this happen often enough to matter, does it follow the same steps each time rather than needing a judgment call, and if the automation gets something wrong, is that mistake easy to catch and fix?
A quick way to test this yourself: if you’ve done a task by hand more than three or four times and it looked basically the same each time, it’s almost certainly a good candidate. Start with the lower-stakes stuff too, internal reports and notifications rather than anything customer-facing or financial, until you’ve built up some trust in how your workflows behave.
This post was just orientation. Next time we’ll actually install n8n, comparing Docker, npm, and n8n Cloud so you’ve got a real instance running before we touch anything else. After that we build a first real workflow together, then work through the data model, conditional logic, error handling, and eventually production self-hosting and AI agent workflows built inside n8n itself.
]]>