Why most personal automations die

I wanted to write about how to connect Google Sheets with Notion through n8n.

Or about how to build a machine that sends you ready-to-approve posts every morning.

But I’ll write about something else.

Recently I’ve built dozens of automations for myself to remove routine, get my time back, and honestly, to feel a bit smarter.

About 80% of them died.

I launched them, tested them, looked at the results, and a few days later I simply stopped opening them.

I figured out why. Three reasons.

1. The hypothesis was wrong

Often we think we’re wasting time on one thing, while the real problem is somewhere else.

I built a system that automatically tailored my CV and cover letter for each job application. It looked great.

Until I noticed that every job application form still has its own questionnaire and you end up pasting everything manually anyway.

The automation solved a problem that didn’t really exist.

2. Compromised UX

“I’ll improve the usability later” is a classic trap.

I built a bot: you say “meeting on Friday at 3 PM,” and it adds the event to your calendar automatically.

I stopped using it because it sometimes made mistakes. I could have fixed them. I didn’t. The bot died.

Todoist implemented the same feature properly, and that’s exactly why people actually use it.

3. It doesn’t scale well

A bot writes texts. The first three are great.

Then you start noticing patterns.

Then you realize they all look the same.

A small simple bot does not produce, out of the box, what you actually want.

Making it truly varied turns into a much bigger project.

Conclusion

First understand where time is actually being lost and why.

Only then build automation.

Otherwise you end up with a very elegant automation solving a problem that never existed.

What usually kills your automations?