If I had a shekel for every time someone asked me in a private message: "Tell me, Tal, what's better: Make, Zapier, or maybe n8n?", I'd be writing this post from the Maldives, not from my home office in Tel Aviv. That question drives me crazy. (Really, my eye twitches when I hear it, and I'm not joking.)
Why? Because it's like asking a chef: "Tell me, will this knife make my food taste good?". The answer is a flat no. The knife is just a tool. If you don't know how to cook, or worse, if your ingredients are rotten, even a thousand-dollar Japanese knife won't save the dish. It'll just help you cut your finger faster.
We live in the age of "shiny object syndrome". People see an ad for a new automation tool, are sure it's the magic fix for every problem in their business, rush to buy a subscription, and then... nothing. The project gets stuck, the process gets tangled, the systems don't talk to each other, and they blame the tool.
The painful truth: automation doesn't fix a broken process, it just speeds it up
This is the number one mistake I see in businesses, and it happens again and again. People take a work process (say, taking in leads or managing a project) that's messy, unclear, full of manual gaps and decisions made "on the fly", and just "throw" automation at it.
The result? Instead of a slow manual mess, you now have a very fast automated mess. The robot (or the bot, whatever you want to call it) does exactly what you told it to do, and if you gave it bad instructions, it will carry them out perfectly, devotedly, and nonstop. (And that, friends, is the sure recipe for disasters on an industrial scale.)
A case from the field: Danny's disaster (not his real name, the embarrassment is real)
A client came to me, let's call him Danny, with a successful e-commerce store. He wanted to "automate everything". He'd heard about Make and decided it was what would save his life. He started building scenarios without end: every order sends an email, opens a Trello card, texts the warehouse worker, updates stock in an external Excel file... In short, one big plate of automation spaghetti.
One day there was a small glitch in his inventory system. The bot didn't know how to handle it. The result? The bot went into an infinite loop and sent that same poor warehouse worker 1,500 text messages within an hour. The warehouse worker almost quit, and Danny paid thousands of shekels for a text message bundle. Why did it happen? Because Danny skipped the most important step: he didn't define what happens when there's an error. He built the "happy path" and forgot that life itself is full of bugs.
So what do you do before you open Zapier (or any other tool)?
Before you touch a single button in an automation tool, before you enter your credit card and before you get excited about the new AI features in n8n, you have to do the boring grunt work nobody likes to do: the spec. It's like laying foundations for a building. It's not sexy, nobody sees it, but if you skip it, the building collapses.
Here are 3 questions you have to answer, in writing (yes, physically, with a pen and paper or a Word document), before you start:
- What's the ideal manual process? Take a pen and paper. Draw the process the way it should happen in a perfect world, if you had a hardworking employee doing everything by hand. What happens after what? Who's responsible for what? What's the trigger that starts the process? If you can't explain it on paper to a 10-year-old, you can't explain it to a robot.
- Where does your data live? Automation is really just moving information from place A to place B. But if your information is a mess, some of it in an Excel file from 2019, some in a notebook, some in your secretary's head, and some on yellow sticky notes on the screen, it won't work. (Garbage in, garbage out. Sound familiar?) Sort out the source before you connect the pipes.
- What's the overarching goal? (And no, "save time" isn't a good enough answer) Be specific. "I want that from the moment a client leaves their details in the form on the website, within 5 minutes they get a first quote on WhatsApp, and their details go into the CRM under the status 'new lead'." Great. Now we have something defined, measurable and clear that can be built.
A pro tip: don't try to eat the whole elephant in one bite
I see this happen a lot, especially with energetic founders (and I relate, I'm one too). They want everything automated, here and now. So they build dozens of processes in parallel. That's a recipe for failure. It's stressful, it creates chaos, and when something breaks, you have no idea where.
Start small. Take one process, relatively simple, that eats up your precious time. Spec it, build it, test it for a week. Make sure it runs smoothly. Only then move on to the next process. That's how you build an automation empire, brick by brick, and not in one big bang that blows up in your face.
The bottom line
Today's tech tools are amazing. Whether it's friendly, visual Make, veteran and reliable Zapier, or powerful, flexible n8n, they all let us do magic we didn't dream of five years ago. But they're only tools.
Your next project won't fail because of a technical limitation in the tool you picked. It will fail because you skipped the planning stage. It will fail because you didn't have the patience to do the grunt work. Stop looking for the perfect tool, and start building the perfect process. The tool will sort itself out.
