n8n for Beginners: Build Your First Automated Workflow Step by Step

Follow this n8n tutorial for beginners to build a scheduled workflow that fetches live data, formats a daily report, saves it, and runs itself.

By Galactic News Editorial · Published 2026-11-06T22:46:52+00:00 · Updated 2026-11-06T22:46:52+00:00

This n8n tutorial for beginners builds one complete workflow from an empty canvas. You will add a schedule trigger, fetch live data, and turn it into a readable report. Then you will save the report and switch the workflow on. The build follows n8n's official quickstart, so every step matches the current product. You need no coding experience. By the end, you will understand the four ideas behind almost every n8n workflow: triggers, nodes, expressions, and publishing.

What you need before you start

You need an n8n instance. The fastest option is n8n Cloud. n8n offers a free trial, and its pricing page states that no credit card is required. The alternative is self-hosting. The Community Edition is a standard, self-hosted version of n8n, and n8n publishes it on GitHub. Pick Cloud for this tutorial. Hosting, updates, and scaling are handled for you, so you can focus on the workflow itself.

The current quickstart builds its example with n8n Assistant. That is a chat tool that turns a written prompt into nodes on the canvas. The docs also include a manual path. This tutorial follows the manual path. Building by hand teaches you how workflows actually work, and that knowledge transfers to every later project.

n8n tutorial for beginners: build a daily weather log

Your workflow will fetch the weather for Berlin every morning. It will turn the raw data into one readable sentence. Then it will save that sentence to a data table, building a weather log over time. Follow these steps in order.

1. Create the workflow. Sign in to your n8n instance. Select Personal in the left menu, then select Create workflow. Name the workflow Weather Log.
2. Add the trigger. Select Add first step, then select On a schedule. Set the Trigger at Hour field to 8am. Rename the node to Every day at 8 AM. The trigger decides when the workflow starts.
3. Fetch the data. Select the plus sign after the trigger. Search for HTTP Request and add the node. Paste https://mockingbird.n8n.io/v1/weather/berlin into the URL field. Mockingbird is a sample-data service n8n provides for tutorials, so no API key is needed.
4. Test this step alone. Press Execute step. You should see the Berlin weather in the output panel. Look for the city, temp_c, and condition fields. You will use those fields in the next step.
5. Shape the report. Add an Edit Fields (Set) node after the HTTP Request node. Add a field named report. Start its value with the words The temperature in. Then drag the city field from the input panel into the value. n8n converts the dropped field into an expression such as {{ $json.city }}. Add the temperature and condition fields the same way.
6. Save the result. Add a Data table node at the end. Point it at a new data table. The node inserts each report as a new row, so the log grows by one row per day.
7. Run the whole workflow. Select Execute workflow. Check that every node completes and that a row appears in the data table. Fix any failing node before you continue.
8. Publish. Select Publish in the canvas header. Publishing activates the schedule. From now on, the workflow runs itself at 8 AM every day.

The four nodes, explained

Each node does one job and passes its result to the next node. That simple pattern is the whole model.

• Schedule Trigger starts the workflow at a set time. Every workflow begins with a trigger of some kind.
• HTTP Request calls a web address and returns the data that address provides. Services that share data this way are called APIs.
• Edit Fields (Set) builds and reshapes fields. Here it turns raw weather values into one readable sentence.
• Data table stores the output inside n8n. Each run adds one row.

Notice what the Edit Fields step demonstrates. Nodes pass structured data, called items, from one step to the next. Expressions let a later node reach into that data and pull out single fields. Once you understand items and expressions, most n8n nodes become variations on a theme you already know.

Triggers come in several forms. The n8n docs pose the starting question this way: should the workflow run on a schedule, such as 8 AM every day? When a user submits a form? When an application like Telegram sends a message? Your answer picks the trigger. Everything after the trigger follows the same node-by-node pattern.

Cloud or self-hosted: what beginners should pick

n8n charges Cloud customers per execution. One execution is a full workflow run, from trigger to finish. n8n puts it plainly on its pricing page: if a workflow does not run, you do not pay. The steps inside a run do not add cost. The Starter plan costs 20 euros per month billed annually, is hosted by n8n, and includes 2,500 workflow executions a month. Pro costs 50 euros per month billed annually. Business costs 667 euros per month billed annually and is self-hosted only. The Community Edition carries no licence cost, because you run it on your own server.

Start on Cloud. Move to self-hosting when you need full control over data or infrastructure. Accept the maintenance work that comes with that control. One more Cloud benefit matters early on: every plan includes unlimited users, according to the pricing comparison table. Your experiments will not trigger per-seat charges. n8n Assistant availability also differs by edition. The docs list it on Cloud Starter and Pro plans, and on self-hosted Community and Business editions once you configure it with your own AI provider account.

Habits that save beginners time

• Execute each node as you add it. A completed node means that step works on its own.
• Rename nodes in plain language. Next month, you will read the canvas, not your memory.
• Test with sample data first. The quickstart uses Mockingbird for exactly this reason.
• Publish only after one full manual run succeeds. A broken schedule repeats the same error every day.
• Open any step later to inspect its input and output. n8n describes its platform the same way: open any step and see what it does.

Once this workflow runs cleanly, change one thing at a time. Swap Berlin for another city. Add a second data source. Send the report to an app instead of a table. Small, working changes teach you more than one large, broken build.

When something goes wrong

Most first-build problems are small and visible. Work through them in this order.

• A node returns an error. Read the output panel of the failing node first. It shows the input the node received and the error it produced.
• The weather fetch fails for your city. Mockingbird covers about 40 European cities, according to the docs. Other cities return an error by design. Pick a covered city for now, or swap in a different data source later.
• The report text looks wrong. Check the expression in the Edit Fields node. Dragging a field inserts it at the cursor position, so misplaced fields are common on a first build.
• Nothing runs at 8 AM. The workflow is probably saved but not published. Publishing is what activates the schedule. A saved workflow and a published workflow are in different states.