Build your first n8n workflow
and keep it running.
A beginner-friendly handbook for marketers and business teams. One example workflow runs through the whole guide, from form submission to classification, logging and routing. The later chapters cover what decides whether automation holds up in daily operation: error handling, credentials, hosting, costs and data protection.
Most workflows are built for the day
everything works.
Building a first automation is straightforward. A form fills a spreadsheet, a message reaches the right person, and the process looks solved. The problems appear later, when a token expires, a field arrives empty, the same record comes in twice or a monthly quota runs out.
The most expensive version of that is silent failure, where a workflow stops running and nobody notices for a week. This handbook treats operation as part of the build, so the automation you publish is one your team can rely on and hand over.
The difference shows up
when something breaks.
Assistants such as ChatGPT or Claude can run scheduled tasks, handle multi-step work and reach email, calendars or files through connectors. The more useful comparison is what happens on a bad day.
For business-critical processes, failure handling and traceability are the durable distinction.
One workflow, extended
chapter by chapter.
A request arrives through a form, gets classified and summarised, is logged in a table and reaches the person responsible. You build it in chapter 1 and keep extending the same workflow through the rest of the handbook.
Five of the six steps are rule-based. Keeping the number of model decisions low is what makes the workflow predictable and cheap to run, which chapter 3 covers in detail.
Ten chapters in three parts.
The first seven chapters need neither programming skills nor your own server. One example, a contact request that is assessed, logged and routed, is extended from chapter to chapter.
Build
Your first working workflow, and the way data actually moves through it.
- Your first workflow in thirty minutes
- How data moves through a workflow
- AI in the workflow, agents compared with simple model calls
- Credentials and access to other systems
Make it reliable
The settings, error paths and structure that turn a demo into something you can operate.
- Settings you should get right once
- Error workflows, retries and the six-step troubleshooting routine
- Traceability, backups and workflow organisation
Decide and operate
The commercial and legal questions that are easier to answer before you scale.
- Cloud or self-hosted, with plans and licence terms explained
- Security, cost control and data protection basics
- When n8n is the wrong tool
Appendices
Reference material you can use while working rather than reading.
- Self-hosting installation with Docker, PostgreSQL and a reverse proxy
- Checklists for setup, publishing and the weekly review
- Glossary and sources
Four situations where
automation is the wrong answer.
A handbook that only explains where a tool helps is incomplete. Chapter 10 is there so you spend effort on the processes that repay it.
Rule of thumb from the handbook: automate after you have run the process manually at least three times and can describe it clearly.
n8n for Beginners
Build your first workflow, understand what happens when it fails, and turn automation into a process you can actually operate. English edition, August 2026.
Download the handbook.
Automate with confidence.
Enter your first name and email and you will receive the English edition as a PDF. Occasional, relevant content on marketing, AI and automation follows. You can unsubscribe at any time.
Your data is safe. I never pass your information on to third parties.
Greye
Fractional CMO & Marketing Strategist
I build and run marketing automation for growing companies, and I teach n8n. The handbook reflects that: the checklists, the failure patterns and the cost logic come from workflows that are in production, not from a feature list.
My work connects marketing strategy with operational execution across B2B SaaS, e-commerce and growth brands. Automation is useful when strategy, process, data and ownership line up, which is why this guide spends as much time on operating a workflow as on building one.
More about me →Where I can support you.
The handbook gives you the model. Which process to automate first, and how it fits your CRM, your data and your team, depends on your setup. These are the areas I work on with clients.
Want a second pair of eyes on your setup?
If your marketing runs on a mix of tools, manual hand-offs and early AI experiments, a conversation is the fastest way to find out which process is worth improving first.