n8n Review 2025: Is It Worth It for Non-Developers?

n8n is a workflow automation platform that’s designed to be developer-friendly. Unlike its competitors, it’s open-source, and it’s highly customizable for people with coding skills.

But what’s n8n like for a non-developer like me? Can you use it without coding skills? And how does it compare to other options like Make and Zapier?

In this review, I'll test n8n from a non-technical user's perspective to unpack whether its technical complexity is worth it for your automation needs.

What is n8n?

n8n (pronounced “n-eight-n”) is an open source workflow automation tool. It lets you connect different apps, APIs, and services, without heavy (or any) coding. But unlike alternatives like Zapier and Make, it’s open source, and you have the option to host it yourself. This gives users full control with no vendor lock-in. As a result, it’s especially popular among developers, privacy-conscious teams, and businesses that want more control than competitors can provide.

Learn more about how n8n compares to other automation platforms in n8n Vs. Make and Make vs. Zapier.

Getting Started With n8n

First, let’s take a look at what setting up looks like with n8n.

Installation and Setup

The first decision you have to make with n8n is whether you want to host the platform yourself or run it in the cloud. The key difference between n8n Cloud and the self-hosted solution lies in who manages the infrastructure and the associated costs.

Self-Hosted Version

To get started with n8n’s self-hosted option, get the latest community version of the code from GitHub.

Cloud Version

Since this review is aimed at business owners and marketers, I’ll cover the cloud version only. There’s also a self-hosted premium (paid) version of the platform that comes with some of the benefits of the cloud version, but the pricing is quite steep.

Managing Your Projects

n8n supports a personal workspace where you can keep workflows private, as well as shared projects (with one shared project available on the Starter plan).

Each project has its own:

• Workflows
• Credentials (app connections)
• Executions (a log of executed automations)

Moving workflows and credentials between projects is easy, and you can even drag and drop workflows.

Creating Your First Workflow

Creating my first simple workflow in n8n was straightforward, but it was also the first thing that annoyed me about the platform. There's no gallery of pre-built templates directly accessible in the cloud app. You have to open a separate tab (through a link in the sidebar) to get access to the directory, then import the workflow template into your cloud workspace.

Once that’s out of the way, if you’ve used Make or Zapier, you’ll feel familiar with n8n's visual workflow builder.

Automation steps, called “nodes” in n8n, are added to the workflow canvas and connected to execute different types of automations.

Types of Nodes

Generally speaking, there are five different types of nodes in n8n:

• Triggers
• Actions
• Flow control nodes or utility nodes
• Human in the loop nodes
• AI agent nodes

Triggers

Triggers start workflows and are always at the beginning of a flow. You can have multiple trigger nodes in a workflow, but you cannot connect a trigger node to another trigger.

Triggers can be various things like an app event (when something happens in a third-party app), a webhook call, a form submission, or a chat message. You can also trigger workflows manually or at a specific time.

Flow Control or Utility Nodes

Confusingly named, this type of node controls the flow of actions and data in the workflow. Think of them as conditional branches or swimming lanes. Popular ones include “Filter,” “If,” and “Merge” nodes.

This is what a branched “If” step looks like in the canvas:

Actions

Action nodes execute actions in workflows. When a trigger is fired, the action follows. n8n supports all kinds of actions, including built-in actions and actions from third-party integrations.

Human in the Loop Nodes

Human-in-the-loop nodes allow you to create manual safeguards, asking a person to review a node output before allowing the workflow to continue. You can see a simple human-in-the-loop workflow here.

AI Agent Nodes

AI agent nodes in n8n are specialized workflow nodes that enable you to build conversational AI assistants within your automation workflows. These nodes work together to create intelligent bots that can understand user input, make decisions, and perform actions. Think of them as a more flexible, smarter version of the other nodes.

If you’re interested in AI assistants, you might also want to read my Synthflow AI Review and my Vapi Voice AI Review.

Setting Up Nodes

The first major UI difference I found in n8n compared to Make or Zapier happens once you start setting up your first node.

You're faced with a three-column layout. The far left and right columns are empty.

The left column is your “Input” data, or the data you’ve received from the previous node that can be used in this node.

The middle column is the node’s main setup panel.

And the right column is the “Output” data that the node produces.

Once you connect and configure the nodes, this UI makes much more sense. In the example below, there are three previous nodes that each send one item of data, and one item of data is generated in the “Output” section.

I’m not sure if I love or hate this type of interface. It’s awkward to drag and drop items across a large panel like this. Also, the full-screen pop-up doesn’t display well on smaller screens/windows. However, I loved that the input column provides easy access to the data from all previous nodes, not just the last connected node.

Working With Data in n8n

An item in n8n is a single unit of data that flows through your workflow. Think of it as one record or data object that gets processed by each node in your automation: a row in a table, a returned record from an API, the response from a form submission, etc.

Each item is a JSON object containing key-value pairs of data. For instance:

{ "name": "John Doe", "email": "john@example.com", "age": 30 }

To use data from items, drag and drop the white data boxes from the “Input” column to the areas in the middle column that allow for data usage. Once the box is dropped, it'll be automatically visualized as a JSON string:

JSON is the backbone of data manipulation in n8n. Every item that flows through your workflow is a JSON object, and understanding how to work with this data is crucial for building effective automations.

This highlights n8n's biggest weakness for non-technical users. Its heavy reliance on JSON data structures and expression syntax can be intimidating for non-developers who aren't comfortable working with code-like formats and curly brace expressions.

Yes, you can get around n8n without code, but you’ll certainly feel much more comfortable if you understand JavaScript. Every piece of information flowing through your workflow exists as JSON objects, which means you'll constantly encounter data structures filled with curly braces, nested properties, and array notation. To modify, filter, or transform this data, you need to use expression syntax like {{ $json.customer.email.split('@')[0] }}. It’s also helpful to know things like {{ $json.field || 'default value' }} to provide fallbacks or {{ $json.customer.address.city }} to access nested data.

Overall, I find working with data in n8n to be more complex than working with data in Make (which is already quite complex). At least Make presents data as clean, labeled fields in dropdown menus and does most of the data transformations in the background.

Testing Your Workflows

The first step when testing and troubleshooting your workflows in n8n is to ensure your nodes work. Each node has an “Execute step” button in the top middle section, which is perfect for granular testing.

When successfully executed, you'll see a green border and a tick next to the node, as well as the output of the node.

To test the whole workflow, you have to click on “Execute workflow.” This will perform a test going through all of the nodes in the workflow.

You can see a thorough log of each executed node at the bottom of the canvas:

Keep in mind that if you have even a single unconfigured or erroneous node in the workflow, you'll get an error when executing the workflow.

Unlike Make, n8n supports sandbox test data, so you don’t need to push real-life data to the workflow every single time to test it. It’s pretty hidden, but there’s a small link - “Insert mock data” - in the Output section of the node. It’s not as robust as Zapier’s sandbox data, but it gets the job done.

To further simplify your testing process, you can pin data items in your nodes by clicking the pin icon in the top right. When a data item is pinned, you don’t have to submit new data each time you test the workflow or the node. Quite convenient!

Overall, troubleshooting n8n workflows is fairly intuitive too. But when a workflow or node fails to execute properly, n8n's error messages can be frustrating. You get a vague "Problem executing workflow" message and have to manually search through the canvas to identify which node is causing the issue.

n8n Key Features

Now let’s take a tour of n8n's key features.

Integrations

n8n boasts an ecosystem of 1,100+ integrations at the time of writing this review, so it’s likely to connect to many of the tools you’ll need.

Almost all integrations provide both trigger and action nodes, allowing workflows to be initiated by external events (like new emails, form submissions, or database changes) and perform operations across your whole app stack. The platform supports major players like Google Workspace, Slack, OpenAI, Anthropic's Claude, Airtable, GitHub, MongoDB, Pipedrive, etc.

In addition, n8n's flexible architecture allows for custom integrations through HTTP requests, webhooks, GraphQL endpoints, and even custom node development.

All that said, n8n’s 1,100 integrations don’t even come close to Make’s 2,800+ or Zapier’s 8,000+ apps. If you're searching for something more exotic or obscure, you might have to look elsewhere or use a workaround like HTTP requests or webhooks.

AI Agents

AI agent nodes in n8n are specialized workflow components that enable you to build conversational AI assistants within your automation workflows. These nodes work together to create intelligent bots that can understand user input, make decisions, and perform actions.

They work differently from standard action nodes, so it might feel confusing when you first try them (at least it did for me). However, once you understand them, they can be pretty powerful, and it quickly became one of my favorite features.

Before you can use the AI agent node, you must add a chat message trigger node. This will start an embeddable or shareable chat widget.

When opened, this node allows you to chat directly in the workflow. This is quite innovative, and something I haven’t seen in other similar platforms.

You can share this chat via a link or embed it in your site. By default, the shareable widget looks like this:

Unfortunately, I couldn’t figure out a way to change the UI of the hosted chat via the node settings, but maybe the embedded one allows you to make changes via CSS.

Once you have the chat node, you have to connect it to an AI agent node.

This is how my workflow looks with Claude and Google Sheets connected:

To test the agent, simply chat with it and ask it to perform an action with the connected tool. For instance, I asked it to list all of the contacts in the table:

You'll get the same results in the chat widget:

AI Capabilities and AI Nodes

Apart from the AI agents, n8n offers comprehensive integration with AI through a collection of more than 70 AI nodes built on top of the LangChain framework. This open-source framework is designed to simplify the development of applications powered by large language models.

These AI nodes work with all the popular LLMs like OpenAI's GPT series, Anthropic's Claude, Google's models, etc., allowing users to generate text, analyze responses, and integrate decision-making capabilities. You can also host your own LLM data for full data privacy.

Most AI nodes are neatly organized in their category and include nodes for LLM chains, information extraction, sentiment analysis, and more.

n8n also supports advanced AI architectures like retrieval-augmented generation (RAG) through vector database integrations. This lets you set up workflows where AI can pull from your own private data and keep its answers accurate and relevant.

Last but not least, the built-in AI assistant will help you learn how to set up nodes and troubleshoot your workflows. I find the assistant on par with Make’s AI helper, but it doesn’t come close to Zapier, where you can create full automations solely by prompting.

Human in the Loop

Human-in-the-loop is a workflow feature in n8n that allows you to pause an automated workflow and wait for human input or approval before proceeding. This creates a checkpoint where a person can review data, make decisions, or provide additional information that the automation needs to continue.

I liked this feature because, unlike some other n8n features, it was created for dummies. The feature supports several messaging apps like Gmail, Microsoft Teams, and Slack.

When setting up the step, you have to define how the person should respond: with a one-click approval, a custom message, or by answering a pre-defined form.

Depending on the app, the person will receive an automated message to respond. Super simple, but it works.

Then, you can connect the “Send Gmail message” node to an “If” node that will branch out the workflow depending on whether you approve the form or not. This is how the If node setup looks:

And this is how my final workflow looks:

Error Triggers

The error trigger feature in n8n provides a robust error-handling mechanism. It automatically activates when any node in a workflow encounters an error. Once enabled, it captures failed executions and routes them to a separate workflow where you can add recovery logic, send notifications, log errors, or retry operations.

The error trigger receives detailed information about the failed execution, including the specific node that caused the error, error messages, and the original workflow data.

Developer-Friendly Features

While this review is focused on n8n's accessibility for non-technical users, it's worth highlighting the platform's robust developer-oriented features that set it apart from other automation tools.

Because it’s open source under a fair-code license, you can see and use all of its code. You can run it yourself on your own servers using tools like Docker or Kubernetes, or use cloud hosting. Developers can also track and manage changes with Git, just like they would in normal software projects.

n8n supports different ways of running tasks, including queue mode (which helps spread out work across multiple machines for bigger projects). On self-hosted setups, you can also add your own custom npm packages. It even supports GraphQL, a modern way of pulling data more efficiently.

All of this makes n8n a strong choice for companies that need advanced, flexible automation. For developers especially, n8n is more customizable, more adaptable, and often more cost-effective than other tools.

n8n Pricing

n8n pricing is based on the number of executions and the plan tier you select.

An execution in n8n is counted when an event or a condition triggers your workflow, and the workflow runs from start to finish. This is certainly more generous than Make’s operations, where each action in your automation, like adding a Google Sheet row or fetching Gmail account data, counts as one operation.

Here’s a look at the supported plans. Note that n8n provides pricing in euros, but you can convert that to other currencies here.

Verdict

n8n shines as a developer-friendly automation platform. If you need self-hosting capabilities, advanced features like custom npm packages and Git-based version control, and the flexibility to build custom integrations and AI-powered workflows, n8n is an exceptional choice.

However, n8n's heavy reliance on JSON data structures, JavaScript expressions, and technical configuration makes it more complex than alternatives like Make and Zapier.

For teams with technical expertise or specific requirements around data sovereignty and security, n8n offers unmatched flexibility and cost-effectiveness. For straightforward business automation without coding, I think Zapier's polished UI, simplified data handling, and huge directory of integrations provide a more accessible path to productivity. While n8n offers impressive technical capabilities and great value, I'd recommend most non-technical users stick with Zapier for their automation needs.

Alternatives to n8n

Make

While n8n excels with its open-source edition and self-hosting features, Make provides sophisticated automations, including visual routers, scenario inputs, and flexible scheduling. The platform handles complex, multi-step workflows exceptionally well through its managed service, though it comes with a steep learning curve and potentially more complex data mapping.

Make offers a more polished, cloud-first experience with 2,500+ native integrations and advanced features like AI-assisted scenario building and automated error handling.

For businesses seeking extensive integrations and enterprise-level features without the technical overhead of self-hosting, Make offers an attractive middle ground between n8n's flexibility and simpler tools like Zapier.

Check out my Make Review to learn more.

Zapier

While n8n excels with its open-source nature, self-hosting capabilities, and developer-friendly approach, Zapier provides an all-in-one managed solution with superior ease of use, sandbox testing environments, and extensive team collaboration features.

Zapier offers a comprehensive AI-powered automation ecosystem with 8,000+ native integrations and advanced features like built-in chatbots, AI agents, databases, and custom interfaces.

The platform's AI-first approach integrates AI throughout the entire interface, from workflow creation to troubleshooting, although this comes at a higher cost.

For businesses prioritizing fast deployment, extensive app connectivity, and comprehensive automation features without technical overhead, Zapier offers a polished, enterprise-ready solution. It trades n8n's flexibility and cost-effectiveness for simplicity and advanced AI features.