Technical 7 min read

Why I Don't Use n8n (And What I Do Instead)

If you've been pitched an AI system recently, there's a good chance you saw n8n in the demo. It demos well. But a compelling demo and a reliable production system are different things; and the distance between them is where businesses get hurt.

If you’ve been pitched an AI system recently, there’s a good chance you saw n8n in the demo. It’s become the go-to tool for AI consultants building automated workflows: connecting data sources, chaining prompts, integrating APIs through a clean visual interface. Nodes connect to nodes, data flows left to right, and it looks like architecture.

It might even have been the moment you thought: “This is real. This person knows what they’re doing.

I don’t use it. Here’s why, and why it matters to you as the person paying for the system.


What n8n Actually Is

To be fair, n8n is a legitimate tool. It’s a visual workflow engine that lets you connect data sources, trigger actions, and pass information between systems (including AI prompts) without writing much code. Think of it as a graphical pipeline builder. You can see how everything connects and wire up integrations relatively quickly.

For certain use cases, that’s genuinely useful. We’ll come back to those.


Why It Looks So Convincing in a Demo

n8n demos well. That’s not a criticism; it’s the product’s greatest strength and the reason consultants reach for it. You see boxes connected by lines, data flowing through each step, and the logic feels transparent. It looks like you’re seeing the system’s brain.

But what you’re actually seeing is the wiring, not the thinking. The visual interface shows you how data moves between steps. It doesn’t show you whether the data is structured well, whether the system will hold up when something unexpected arrives, or what happens when the person who built it moves on.

This is the gap that matters. A compelling demo and a reliable production system are different things. The distance between them is where businesses get hurt.


The Handover Problem

Here’s the scenario that plays out more often than it should.

A consultant builds you an n8n-based system. The demo is impressive. The workflows run. You sign off, pay the invoice, and the engagement ends.

Three months later, an API credential expires. Or Twilio changes their webhook format. Or Google updates their Calendar API. The workflow that looked so clear in the demo is now a red node with an error message nobody in your business understands. You call the consultant (maybe they’re available, maybe they’ve moved on). Either way, you’re dependent on someone external to keep your own system running.

This isn’t a rare edge case. APIs change. Credentials rotate. Platforms update. In a code-based system, these are routine maintenance tasks that any competent developer can diagnose and fix. In an n8n workflow, they require someone who understands both the platform and the specific way your workflows were built. That’s a smaller pool of people, and you don’t control it.

The visual interface that made the system easy to understand in a meeting makes it no easier to fix when something breaks. The nodes and connectors that looked like clarity become a dependency you didn’t sign up for.


The Deeper Problem: It Can Mask Weak Data Thinking

Many consultants reach for n8n before they’ve sorted out the data model. The visual interface creates an illusion of architecture. You can wire together a sophisticated-looking workflow without ever asking the hard question: what does my data actually need to look like?

The nodes and connectors become a substitute for that discipline. You end up with workflows that are hard to debug, hard to scale, and hard to hand off. They work until they don’t. And when they don’t, diagnosing the problem means tracing data through a visual pipeline that was designed to look good, not to fail gracefully.

n8n doesn’t solve your data problem. It can paper over it.


The Hidden Operational Cost

Beyond the handover risk, there’s a running cost that rarely gets discussed in the sales conversation.

In a typical small or mid-sized business, the n8n platform needs to be installed and maintained (either self-hosted for your responsibility, or on a cloud subscription with ongoing cost). Every API connection needs to be credentialed and maintained separately. Updates to the platform itself can break existing workflows. And all of this accumulates quietly in the background while you’re focused on running your actual business.

Good consulting should leave you more capable and more independent. Not tethered to a third-party platform you didn’t choose and don’t fully understand.


Opaque to the Tools That Matter Most

There is another cost that is easy to miss: visual workflow configurations are opaque to AI itself.

We wrote recently about why codebases need to be legible to AI. The systems that AI can understand, maintain, and extend are the ones built with explicit structure and clear interfaces. Code is legible. A well-structured data model is legible. An n8n workflow exported as JSON is a proprietary format that no AI coding tool can reason about, debug, or extend without deep platform-specific knowledge.

This matters because AI-assisted development is not a future consideration; it is how software is maintained now. A code-based system can be understood, modified, and improved by AI tools. An n8n workflow cannot. You are locking your business logic into a format that the most powerful development tools available cannot work with directly.


The Better Starting Point: Your Data Model

I start with a different question: what does this data actually need to look like?

Before writing a line of code, I design clean, discrete, well-structured data models for each domain I’m working with: a clear schema for each part of the business. When that model is right, the pipeline logic becomes almost self-evident. The structure tells you what transformations are needed and how the system should behave.

The data is the architecture. The pipeline just follows.

This might sound abstract, but here’s what it means in practice. When you need to change something (add a new field, connect a new service, generate a new report), the change is straightforward because the foundation is sound. You’re not re-routing a visual pipeline and hoping nothing downstream breaks. You’re updating a model that everything else is built on.


Layer Sophistication Without Losing Control

Once the data model is solid, you can stack increasingly powerful capabilities on top: document analysis, photo and video processing, image generation, complex reasoning chains. All are coded robustly and all feed back into the same expanding data structure. The model grows and enriches over time, staying coherent because the foundation was right from the start.

This answers the obvious objection: can you do complex things without n8n? Yes. And you can do them more reliably, because your foundation holds.


Robustness You Actually Own

Visual workflows are only as robust as the data flowing through them. When an API changes or a feed arrives malformed, workflows can fall over in ways that are genuinely difficult to diagnose. n8n does offer basic retries and logs; but you don’t own the underlying control plane the way you do with code. With code you get proper error handling, input validation, circuit breakers, and logging you actually control. When something breaks (and it will), the system fails gracefully and tells you exactly why.

That’s the difference between a prototype and a production system. And it’s the difference between a system that needs its builder on speed-dial and one that any competent developer can maintain.


The Real Payoff: Portable IP

Here’s the part that matters most to a business owner thinking long-term.

Your data becomes the asset. n8n workflows are exportable, but they’re coupled to how n8n thinks about execution. Move to a different platform and you’re rebuilding. A clean data model is portable to any stack. Platform-independent. Extractable in a single clean pull. If you need to swap providers, change your processing layer, or scale the system, the data comes with you.

With n8n, you’re building workflows tied to a platform. With clean data models and robust code, you’re building business infrastructure that you own outright.


Where n8n Actually Earns Its Place

n8n is genuinely useful as a bridge tool: integrating messy legacy systems, wiring together disparate data sources as a pragmatic first step before the clean architecture is in place. That’s a legitimate and specific use case.

But it’s a bridge, not a destination.


If You’ve Already Been Sold an n8n System

If you’re already running n8n workflows, you don’t have to throw them away overnight. Start by asking: what does the underlying data actually look like? Is it structured, consistent, and portable? Or is it scattered across workflow nodes with no coherent schema?

Then progressively migrate your most brittle workflows into code, starting with the ones that break most often. Let n8n continue handling legacy integrations at the edges while the clean architecture is built alongside it. Over time, n8n’s role shrinks to the periphery and your core system becomes robust, portable, and genuinely yours.


The Question Worth Asking

Next time someone shows you a workflow tool with an impressive visual interface, ask one question:

What happens to this system when you’re not here to maintain it?

If the answer involves you calling them back, the architecture isn’t serving you. It’s serving them.


Perth AI Consulting designs AI systems on solid data foundations: robust, portable, and built to outlast the consultant who delivered them. Start with a conversation.

Published 12 March 2026

Perth AI Consulting delivers AI opportunity analysis for small and medium businesses. Start with a conversation.

Written with Claude, Perplexity, and Grok. Directed and edited by Perth AI Consulting.

More from Thinking

Building 9 min read

How We Built On-Device De-Identification So AI Never Sees Real Names

Most AI privacy is a policy. Ours is architecture. We run a named entity recognition model inside the browser to strip identifying information before it ever leaves the device. Here is how it works, what we tested, and where it applies.

Building 8 min read

Your Practice Needs an AML/CTF Program by July 1. Here's What That Actually Looks Like.

AUSTRAC's Tranche 2 reforms hit accountants, real estate agents and settlement agents on 1 July 2026. We built a complete compliance program for a small practice in three days. Here's the process, the output and the boundaries.

Technical 7 min read

Your Agency's Clients Are About to Ask Why This Costs So Much

A solo consultant just built in two weeks what your agency quoted eight for. The client doesn't understand AI yet; but they will. The agencies that survive aren't the ones that cut costs. They're the ones that change what they sell.

Adoption 6 min read

What Do You Love Doing? What Do You Hate Doing?

Most AI rollouts fail the same way. Leadership announces efficiency. Staff hear replacement. A developer at a recent peer group meeting offered a reframe that changes everything; the psychology of why it works tells you how to deploy AI without destroying trust.

Technical 10 min read

Your Codebase Was Not Built for AI. That's the Actual Problem.

Amazon's mandatory meeting about AI breaking production isn't an AI tools story. It's an architecture story. The codebases AI is being pointed at were never designed to be understood by anything other than the humans who built them.

Adoption 4 min read

Your Team Has AI Licences. You Don't Have an AI System.

Fifteen people, fifteen separate AI accounts, no shared context. The problem isn't the tool; it's the architecture around it. Here's what fixing it looks like.

Building 7 min read

Your $2,000 Day Starts the Night Before: Our System Keeps You on the Tools, Not on the Phone

Your route is optimised overnight. Your customers are notified automatically. When something changes mid-day, every affected customer gets told without you picking up the phone. A tradie scheduling system that protects your daily rate.

Evaluation 4 min read

The Fastest Way for an Executive to Get Across AI

AI is moving faster than any executive can track. The alternatives: learning it yourself, sitting through vendor pitches, hiring a consultant who arrives with a hammer, all waste your scarcest resource. There is a faster way.

Building 6 min read

Your IT Department Will Take 18 Months. You Need This Working by Next Quarter.

Senior leaders often know exactly what they need built. The gap isn't technical; it's time. A prototype approach gets the tool working now and gives IT a validated blueprint to build from later.

Adoption 4 min read

What If You Had Perfect Memory Across Every Client?

Any practice managing dozens of ongoing client relationships captures more than it can recall. AI gives practitioners perfect memory across every interaction, so preparation time becomes thinking time, not retrieval time.

Building 8 min read

We Built an AI Invoice Verifier. Here's Where It Hits a Wall.

We built an AI invoice verifier and watched a fake beat a real invoice. Here's why document analysis alone cannot stop invoice fraud; the five layers of detection that most businesses never reach.

Building 5 min read

How to Build an AI Chatbot That Doesn't Lie to Your Customers

Woolworths deliberately scripted its AI to talk about its mother. The business fix is simple: be honest about the bot. The technical fix is harder: architecture that prevents fabrication by design, not by hope.

Technical 9 min read

Why AI Safety Features Are Load-Bearing Architecture, Not Political Decoration

The 'woke AI' label came from real failures; but they were engineering failures, not safety failures. Understanding the difference matters for every organisation deploying AI where errors have consequences.

Adoption 3 min read

Woolworths' AI Told a Customer It Had a Mother. That's a Problem.

Woolworths' AI assistant Olive was deliberately scripted to talk about its mother and uncle during customer calls. When callers realised they were talking to an AI pretending to be human, trust broke instantly.

Evaluation 4 min read

Google Is No Longer the Only Way Your Customers Find You

People are using ChatGPT, Perplexity, and Gemini to find businesses. The sites that get cited are structured differently to the sites that rank on Google. Most businesses are optimising for one and invisible to the other.

Evaluation 4 min read

Two Types of AI Assessment: And How to Know Which One You Need

Most businesses considering AI face the same question: where do we start? The answer depends on whether you need to find the opportunities or reclaim the time. Two assessments, two perspectives, one goal.

Evaluation 4 min read

The Personal Workflow Analysis: What Watching a Real Workday Reveals About Automation

When asked how they spend their day, most people describe the work they value, not the work that consumes their time. Recording a typical workday closes that gap, revealing automation opportunities no interview could surface.

Evaluation 4 min read

What a Good AI Audit Actually Delivers

A useful AI audit produces two things: a written report with specific, costed recommendations and a working prototype you can test. Not a slide deck. Not a proposal for more work.

Evaluation 4 min read

Your Website Looked Great Five Years Ago. Now It's Costing You Customers.

The signals that used to build trust online (polished design, stock imagery, aggressive calls to action) now trigger scepticism. Most businesses don't realise their digital presence is working against them.

Evaluation 4 min read

AI Audit That Starts With Your Business

Most AI consultants arrive with a toolkit and look for places to use it. An operations-first audit starts with how your business actually runs, and only recommends AI where the evidence says it will work.

Building 6 min read

What Production AI Teaches You That Demos Never Will

The gap between AI that works in a demo and AI that works in your business is where the useful lessons live. Architecture, framing, privacy, and adoption; the patterns are the same every time.

Adoption 6 min read

The Psychology of Why Your Team Won't Use AI

You buy the tool, run the demo, and three months later nobody is using it. The reason is not the technology; it is five predictable psychological barriers. Each one has a specific strategy that overcomes it.

Technical 4 min read

Stop Telling AI What NOT to Do: The Positive Framing Revolution

Most businesses get poor results from AI because they instruct it with constraints and prohibitions. Switching from negative framing to positive framing transforms output quality, and the principle comes from psychology, not computer science.

Building 5 min read

How We Turned Generic AI Into a Specialist: And What That Means for Your Business

Most businesses get mediocre AI output and blame the model. The fix is almost never a better model; it's a better architecture. Three structural changes that transform AI from 'fine' to 'actually useful.'

Evaluation 5 min read

Your Business Has 9 Customer Touchpoints. AI Can Fix the 6 You're Dropping.

You are spending money to get customers to your door. Then you are losing them because you cannot personally follow up with every lead, nurture every client, and ask for every review. AI can handle the touchpoints you are dropping: quietly, consistently, and at scale.

Technical 5 min read

What Happens to Your Data When You Press 'Send' on an AI Tool

Most businesses are sending customer data, financials, and internal documents to AI tools without understanding what happens during processing. The spectrum of AI privacy protection is wider than you think; recent research shows that even purpose-built security can have structural flaws.