Tools & Resources
Quick Takes: Things I Automated This Quarter
The automation wins that actually stuck. No time spent building systems that don't work.
Most automations fail because people build systems that solve problems that don’t actually exist. This quarter I automated four things, and exactly three of them stuck. The one that didn’t? I deleted it without guilt. Here’s what worked.
Invoice sending reminders. I have three recurring clients on monthly invoicing. Instead of writing the same email three times or maintaining an embarrassing spreadsheet, I set up a Gmail label filter that pops a reminder when the due date hits. Not full automation, just the trigger. I still write the actual email (it takes two minutes, and it matters that they’re personal). But I never forget, and I never miss a day. This works because the pattern is rigid and the stakes are low enough that a reminder is better than a full system.
Receipts to a folder. Every purchase gets forwarded to a specific email address that automatically files it into a Drive folder by date. This one is pure setup-and-forget. I haven’t thought about it since implementation. It works because there’s no decision-making involved. Receipt comes in, gets organized. That’s it. No edge cases, no exceptions.
Podcast episode routing. New episodes from three shows I follow get dropped into a specific Slack channel automatically. Sounds trivial until you realize I now see them without subscribing to yet another newsletter or checking five different apps. The automation doesn’t save hours, but it removes friction. That matters more than you’d think.
Social post drafting. This is the one I killed. I tried to automate creating social posts from my newsletter. The theory: write once, post everywhere. The reality: the automation created drafts that were worse than worthless because I’d spend 15 minutes revising them anyway. I spent three hours building it and ripped it out after two weeks. No shame. Better to know it doesn’t work than to pretend it’s helping.
The pattern I notice: automation that works removes repetition and requires zero decision-making. Automation that fails tries to handle variability or decision logic. I also notice that the best automations are invisible. They do something and I never think about them again. The ones I’m constantly tinkering with are liabilities masquerading as wins.
Most automation content tells you to automate everything. I’m telling you to automate almost nothing. Pick the three things that happen on the exact same schedule, in the exact same way, and won’t break if they’re slightly wrong. Automate those. Ignore the rest.
If you’re drowning in tools and systems already, read stop collecting tools, start wiring systems first. It’ll save you from building automations you don’t need. And if you want to know which automation tools are actually worth your time, check out automation tools that actually save time for the honest reviews.