AI agents vs. AI automations
Why 90% of marketers don't need AI agents (yet)
I've been building AI automations for two years now, and although I’m constantly playing around with AI agent builders, most of you can probably skip AI agents (for now).
I see these incredibly complex n8n workflows that include AI agents shared all over LinkedIn and YouTube, and they're so unnecessarily complicated for what most marketers actually need to do.
Why most AI agents are overkill for marketing
Look, my perspective is pretty straightforward - I focus on marketing workflows that actually move the needle, not just cool tech demos that might have a practical use case sometime in two years.
As marketers, we need to do simple stuff.
Create content, post content, handle some admin tasks.
The quality of the content has to be great (now even more than ever before), but this doesn't come from fancy AI agents - it comes from good inputs, solid prompts, and human strategy behind the automations/agents.
Even when you do find a use case for an agent instead of automation, you don't need some crazy complex n8n setup.
You need something simple that actually works (and don’t worry, I will share all my best AI agent templates with you once I have validated their actual use cases)
My go-to tools (that actually work)
As of right now, I still find AI automations more valuable than agents most of the time. My toolkit, as you know:
Make.com - Super visual, intuitive for non-techies like me but still allows to build almost everything you can think of. The AI agent features are not that easy to use, though, so I prefer it just for automation workflows.
Airtable - Like Google Sheets on steroids, but so much better for storing data (outputs from AI workflows) and triggering automations. I can also build forms, interfaces, dashboards, and simple built-in automations with it.
Apify - For any type of web scraping that I need to do. I use it for scraping Youtube and TikTok videos, Facebook Group posts and it fits into the Make.com workflows so well.
These tools let me build both complex and simple workflows without needing a tech degree.
N8N: the good, the bad and the ugly
Don't get me wrong - n8n is a good tool. But it's overkill for most non-techie marketers.
Using n8n actually makes sense if:
You're technical and understand how to troubleshoot it when something breaks
You have custom internal tools that need integration and none of the other simpler tools (Make, Zapier) don’t work for your use case
You absolutely need self-hosting for privacy reasons (this is the biggest selling point of n8n)
For everyone else, it's unnecessarily complex. The other tools (Make and Zapier for automations, Lindy AI and Relevance AI for agents) do the same thing with way less complexity.
Why I like Lindy AI as a simpler alternative (for now)
I've been testing Lindy AI, and I think it's heading in the right direction. You can build simple AI agents with natural language prompts - so there’s no technical knowledge required.
The only downside is that it's a bit pricier (at $50/month for 5,000 credits, and those credits burn through quickly). But if it saves you critical time, it's definitely worth it.
4 critical questions to ask yourself before starting to build AI agents or automations
I want to share some of my best practices for how I approach automation/agent building.
Before you dive into any AI automation or agent project, ask yourself:
Do I already know how to do this task manually? You need to know what good output looks like to automate it successfully.
Does this actually save me time? Sometimes setting up automation creates more problems than it solves.
Am I automating something I do regularly? Don't build something you'll use twice a year.
Have I properly evaluated different tools or just believed the hype? Just because n8n looks complex and cool doesn't mean it's better. And just because someone on Youtube says they built an AI agent with it that makes them $10k/mo on autopilot, doesn’t make it true.
The bottom line
Things are getting simpler, not more complicated. That's the beauty of AI. Don't fall for the hype of overly complex setups when simple solutions work just as well.
The goal isn't to build the coolest AI setup. It's to build something that actually works for your business.