n8n consultant: what to look for, and when you need one

What n8n is, and when it beats Zapier or Make

n8n is a node-based automation tool. You connect triggers and actions on a canvas, and where the built-in nodes fall short you drop in a code step. The licence is source-available rather than fully open source, which in practice means you can self-host it for your own business without paying, and you pay if you use their cloud or need enterprise features.

The reason people move to n8n is usually one of three things. Data control, because a self-hosted instance keeps customer records on infrastructure you own rather than passing through a third-party automation service. Cost shape, because n8n cloud charges by workflow execution rather than by every individual step, so a fifteen-step workflow does not cost fifteen times a three-step one. Or capability, because the workflow needs loops, retries, real conditional logic or a bit of code, and the simpler tools start fighting back.

If your workflows are two or three steps between mainstream apps and nobody in the business writes code, n8n is the wrong tool and an n8n consultant is the wrong hire. Say that out loud before you go looking for one.

What good n8n work looks like

Anyone can drag nodes onto a canvas until data moves. The difference between a workflow that survives a year and one that fails silently in week three is all in the parts nobody demos.

• Error handling on every external call. APIs time out and rate-limit. Each step that talks to another system needs a defined retry policy and a defined behaviour when the retries run out.
• Alerting you will actually see. A failed run should land somewhere a human looks daily, usually a Slack channel or an email to a shared inbox. Silent failure is the most expensive failure mode in automation.
• Credentials and accounts in your name. The n8n instance, the API keys and the connected accounts belong to your business, not the consultant. Check this before the first invoice, not after the relationship ends.
• Documentation a non-builder can read. One page per workflow: what triggers it, what it touches, what it writes, what to do when it breaks, who to call.
• Sensible naming and version control. Workflows named for what they do, not workflow 7 copy final. Exported JSON kept in a repository so you can see what changed and roll back.
• A test path. A way to run the workflow against sample data without emailing a real customer or writing to your live accounting file.
• Handover. A recorded walkthrough and an hour with whoever will own it internally. If nobody in your business can open the workflow and understand it, you have bought a dependency.

How to brief an n8n consultant

Brief the outcome and the constraints, not the implementation. Write one paragraph per workflow: what starts it, what has to happen, which systems it touches, what the output looks like, and how often it runs. Add your volumes, because a workflow running fifty times a day is a different build from one running twice.

State the non-negotiables up front. Whether it must be self-hosted, where data may live, who owns the accounts, what documentation you expect, and who inside your business will maintain it afterwards. Those four lines change the quote more than anything else in the brief, and leaving them out is how you end up with a workflow you cannot move.

Ask for the build to be quoted as a fixed price per workflow, with support priced separately and explicitly. A single blended monthly fee makes it impossible to tell what you are paying for or what happens if you stop.

What it costs and how people charge

AI Consulting is A$1,195 per month + GST, with a call each fortnight, clarity on what to build in Claude or ChatGPT and practical guidance on how to do it. No lock-in.

Custom development is A$150/hour + GST, with a separate project scope, estimated hours and terms before work starts.

We agree the work, estimated hours and engagement terms before starting.

Questions to ask before you sign

1. Who owns the n8n instance, the API keys and the connected accounts when this is finished?
2. How will I know when a workflow fails, and where does that alert go?
3. What documentation do I get, and can someone non-technical follow it?
4. Is the build a fixed price? What is explicitly out of scope?
5. What happens when one of these APIs changes? Is that included in support or is it new work?
6. Can you show me a workflow you have built, including its error handling, not just the happy path?
7. If we part ways, what exactly do I keep, and can I export it?

If those answers are vague, the price is not the problem. Before you hire anyone, it is worth knowing whether the workflow deserves to exist. AI Consulting is A$1,195 per month + GST, with a call each fortnight, clarity on what to build in Claude or ChatGPT and practical guidance on how to do it. No lock-in. Custom development is A$150/hour + GST, with a separate project scope, estimated hours and terms before work starts.