Your first real n8n flow: automation for people who don't code

Level 3 explains what automation is and why n8n is often the better choice than Zapier. What has been missing is the moment where you actually build something. We make up for that here. At the end a workflow runs that does something on its own, triggered by an event, without you sitting there. You do not need programming knowledge for it. You only have to understand three building blocks and connect them. As an example I take a classic that is immediately useful for freelancers and career changers. A new event comes in, n8n processes it, you get a notification.

Step 1, decide cloud or self-hosting

There are two ways to start with n8n, and the decision hangs on one question only. Do you want to deal with a server or not.

n8n Cloud costs 24 USD a month on the starter tier for 2,500 executions, Pro sits at 60 USD for 10,000. In exchange n8n takes care of updates, SSL and backups. Self-hosted, the software is free and runs without an execution limit on your own server, but costs you time to learn. n8n itself recommends cloud if you are not experienced with running servers. For your very first flow I would take the free cloud trial, so you can concentrate on building rather than on Docker.

Step 2, the one command if you do self-host

If you do want to try it locally and already have Docker installed, two lines in the terminal are enough. This is the official command from the n8n docs, as of stable version 2.23.4.

docker volume create n8n_data

docker run -it --rm \
  --name n8n \
  -p 5678:5678 \
  -e GENERIC_TIMEZONE="Europe/Berlin" \
  -e TZ="Europe/Berlin" \
  -e N8N_ENFORCE_SETTINGS_FILE_PERMISSIONS=true \
  -e N8N_RUNNERS_ENABLED=true \
  -v n8n_data:/home/node/.n8n \
  docker.n8n.io/n8nio/n8n

Then you open http://localhost:5678 in the browser. Do set the timezone to yours, otherwise time-based flows fire at the wrong hour. That is the most common beginner trap and I walked into it myself when a daily report arrived at one in the morning instead of nine.

Step 3, the mental model, three building blocks

Before you click, pause briefly. An n8n flow always consists of the same parts. A trigger starts the flow. A node does something with the data. A connection, the little line between two nodes, passes the result on.

That is all it is. Every complicated workflow you see later is just a longer chain of exactly these three things. Once you have understood that, you can read anyone else's flow.

Step 4, build the trigger

The trigger is the first node and the only one every flow strictly needs. To start with, two trigger types are the most useful. A schedule trigger that fires on a timetable, say every morning at nine. Or a webhook trigger that fires as soon as another application sends a message to a URL.

If you are not sure what a webhook even is, read the lesson on it first, that saves you a lot of guessing. For your first time take the schedule trigger, it is the easiest to test because you can fire it by hand.

Step 5, build the action

Now you attach a second node to the trigger that does the actual work. That can be an HTTP request that fetches data from a website or API. Or an app node that talks directly to a service you use, a spreadsheet or a CRM for instance.

Connect the two nodes with a line. That line is not decoration, it tells n8n that the result of the trigger should flow into the action. A node without an incoming connection is simply skipped during the run, another typical beginner mistake.

Step 6, add the notification

The third node sends you the result. An email, a Telegram message, an entry in a chat. That is the part that makes the flow tangible in your day, because the result reaches you without you having to look.

For the node to be able to send, you need credentials once, so access data for the respective service. You store them in n8n once and can reuse them in every flow afterwards. Set them up properly, forgotten or wrong credentials are the number one reason why a finished-looking flow still sends nothing.

Step 7, test every step individually

This is the part where n8n is better than most competitors. You run the flow by hand and see for every node what goes in and what comes out. Input on the left, output on the right, in real data.

Click through the nodes and look at the outputs. If the trigger delivers empty data, nothing arrives at the next node and the chain breaks. You need this view of input and output every time something does not run. Instead of guessing, you see exactly where the data disappears.

Step 8, the difference between test and active

An important point that confuses many people. As long as you are testing in the editor, a webhook trigger uses a test URL that only listens once. Only when you switch the flow to active does it get its production URL, which is permanently reachable.

So if you test your webhook from outside and nothing happens, first check whether you are using the right one of the two URLs and whether the flow is active at all. Otherwise that costs you half an hour of debugging in completely the wrong place.

Step 9, let it run and watch

Switch the flow to active and wait for the first real run. n8n logs every execution with a timestamp and status. After the first day, go through the list once and see whether everything ran through or whether there were errors.

This is exactly where the advantage over a self-written script lies. You do not have to monitor anything, the history is right in front of you. A failed run is no drama, you open it, see the node that stumbled, and fix exactly that one.

Step 10, common mistakes at a glance

Four things catch almost everyone on their first flow. The timezone is wrong, so schedule triggers fire at the wrong time. A connection between two nodes is missing, so a node gets skipped. The credentials are not set or have expired, so the last node sends nothing. And the flow is not switched to active, so nothing at all happens on the real event.

If your flow does not do what it should, go through these four in order before you suspect anything else. In nine out of ten cases it is one of those four points.

What next

If you want the basic model to settle first, go back to What is automation. For the honest decision on when n8n really beats Zapier, n8n vs Zapier is required reading. And if you want to properly understand the webhook trigger from step 4, read Understanding webhooks, after that you can build the trigger blindfolded.

Source

• Official Docker installation and current version: https://docs.n8n.io/hosting/installation/docker/
• n8n cloud pricing 2026 (starter 24 USD, Pro 60 USD): https://n8n.io/pricing/