n8n Tutorial for Beginners: The Most Complete Guide

n8n looks intimidating the first time you open it: a blank canvas, hundreds of nodes, words like “items” and “expressions”. This n8n tutorial for beginners breaks it down into the concepts you actually need, explains cloud vs self-hosted, introduces the AI Agent node and walks you through a first real workflow. Read it, then open n8n and build. Learning by doing beats reading ten more tutorials.

n8n is a workflow automation platform that connects apps, APIs and AI models through a visual editor, so tasks like moving leads, sending notifications or answering with AI run on their own. Its code is available, you can host it yourself, and you can add custom code when the visual nodes are not enough.

What do you need to know before this n8n tutorial for beginners?

Nothing technical. You need a clear task to automate and the accounts of the apps you want to connect. Knowing basic JSON helps, and you will pick it up along the way.

Before you start, write the task in one sentence: “When a new lead fills my form, save it in a spreadsheet and email me.” That sentence is your workflow.

Cloud or self-hosted: which n8n should you use?

Use n8n Cloud if you want to start in minutes and do not want to manage servers. Self-host if you want more control, your data on your own server or no execution-based billing, and you are fine handling updates and security.

• n8n Cloud: hosted by n8n, paid plans billed mainly by workflow executions, with a trial to test it.
• Self-hosted Community Edition: free software you run with Docker or npm on your computer or a VPS. You pay for the server and you are responsible for backups, security and updates.

Plans change, so check the current pricing page. For learning, either works.

The 6 core concepts of n8n

Everything in n8n is built from these six ideas. Understand them and the rest is practice.

1. Workflow

A workflow is the full automation on the canvas: a chain of connected nodes that starts with a trigger and runs a sequence of steps.

2. Nodes

A node is one step. Some connect to apps (Gmail, Google Sheets, Slack, Telegram, HTTP Request for any API). Others transform data (Edit Fields, Filter, Merge) or control logic (If, Switch, Loop).

3. Triggers

A trigger node starts the workflow. The most common are:

• Manual trigger: you click a button. Perfect for testing.
• Schedule trigger: runs every hour, day or week.
• Webhook: runs when another app sends data to a URL.
• App triggers: a new email, a new row, a new form submission.
• Chat trigger: starts a conversation with an AI agent.

4. Items and data

Data flows between nodes as a list of items, each one a JSON object. If a trigger brings 5 new rows, the next node usually runs once for each of the 5 items. Seeing the input and output panels of each node is the fastest way to understand what is happening.

5. Expressions

Expressions let you use data from previous nodes inside a field. They are written between double curly braces, for example {{ $json.email }} to insert the email of the current item. You can drag fields from the input panel and n8n writes the expression for you.

6. Credentials and executions

Credentials store the connection to each app (API keys, logins) securely, so you set them once. Executions are the history of every run: what came in, what came out and where it failed. When something breaks, the execution log is the first place to look.

Your first workflow, step by step

Let us build a lead workflow: a form submission gets saved in Google Sheets, AI classifies it and you get an email for hot leads.

1. Create a new workflow and add an n8n Form trigger (or a Webhook if your form is elsewhere) with fields for name, email and message.
2. Add a Google Sheets node, connect your credentials and choose “Append row”. Map name, email and message with expressions.
3. Add an AI step: a node that uses a chat model (Claude, ChatGPT or another) with a prompt like “Classify this message as hot, warm or cold based on buying intent. Answer with one word.” Pass the message with an expression.
4. Add an If node: if the classification equals “hot”, continue on the true branch.
5. Add a Gmail or email node on the true branch that sends you the lead details.
6. Test it: fill the form with a fake hot lead and a fake cold one. Check each node’s output.
7. Publish the workflow so it runs on its own (older versions call this “activate”).

That is a real business workflow. Next step: add an automatic reply to the lead. Our guide on lead follow-up automation shows what a good follow-up sequence looks like.

What is the AI Agent node in n8n?

The AI Agent node lets an AI model decide which steps to take to complete a task, using tools you give it. Instead of a fixed path, the agent reads the request, chooses a tool, uses the result and answers.

An agent in n8n combines:

• A chat model: the AI that thinks and writes.
• Memory: so it remembers the conversation.
• Tools: other nodes or workflows it can call, like searching a sheet, checking a calendar or calling an API.

Example: a chat trigger plus an agent that can look up your price list and book a call. Powerful, but start simple. Give agents few tools, clear instructions and limits, and test them with tricky questions before putting them in front of customers.

Beginner mistakes to avoid

• Building everything at once: build one node, test it, then add the next.
• Not reading the output: most errors are a field with a different name than you expected.
• Forgetting error handling: set up an error workflow or notifications so failures do not go silent.
• Exposing a self-hosted instance carelessly: use HTTPS, strong passwords and keep it updated.
• Trusting AI output blindly: validate what the model returns before using it in decisions.

What should you automate next with n8n?

Pick the task that costs you the most hours. If you are not sure which one, our guide to calculating the cost of manual work helps you rank them. Good next projects:

• A weekly report that pulls numbers from several apps into one email.
• Saving email attachments to a folder and logging them in a sheet.
• Notifying your team in Slack or Telegram when a sale comes in.
• Summarizing long emails or documents with AI.

For more ideas, see 5 business tasks AI can do for you.

Go build

You will break things. That is the training. Every failed execution teaches you more than an hour of videos. Push until something works, fix what breaks, and come back to build the next one. Every hour of busywork you remove is an hour you can give to clients, to your craft, to the people your skills can actually help.

And if you would rather have someone build the critical workflows while you learn the basics, at King Studio AI we design and set up n8n automations and AI chatbots for businesses.