Inbox Zero is a nice idea, but for founders it’s a myth. Between support requests, vendor spam, and the occasional real prospect, your inbox is a full-time job you never asked for. I’ve been there — I spent the first year of my startup manually sorting 150 emails a day.
What changed everything was not a new productivity app. It was automation. Specifically, three scripts that now handle the boring parts of email triage while I focus on actual work.
First, I wrote a simple Python script that connects to Gmail’s API and automatically labels emails based on sender domain and keywords. Anything from ‘unsubscribe’ goes straight to a ‘Newsletters’ folder, skipped from my inbox. That cut my visible emails by 40% immediately.
Second, I added a rule that flags emails from known clients with a high-priority label and moves them to a ‘Needs Reply’ folder. This ensures I never miss a client email buried in the noise. It uses a simple CSV list of client domains, which I update once a month.
Third, for the truly repetitive emails — like ‘I forgot my password’ or ‘How do I upgrade?’ — I implemented an auto-reply script that sends a helpful response with links to relevant docs. Support tickets dropped by 50% because most people don’t need a human response; they just need the right link.
The total setup time was about 3 hours, but it saves me at least 10 hours a week. If you’re drowning in email, start with the label script — it’s the highest ROI.
In this post, I’ll walk you through the exact code for each script and how to schedule them with cron. No fluff, just working examples.
But first, let’s talk about why most ’email automation’ advice fails. Most people try to automate sending emails, not sorting them. That’s backwards. The real win is in triage: deciding what needs your attention and what doesn’t. Once you’ve done that, the rest follows.
Another mistake is over-engineering. You don’t need natural language processing or machine learning. A few well-crafted rules handle 90% of the job. Start with the rule-based approach and add complexity only if you need it.
One caveat: this only works if you use labels aggressively. Don’t be afraid to auto-archive emails that are more than 30 days old and unread — if they’re important, someone will follow up.
Finally, test your scripts on a dummy account first. There’s nothing worse than accidentally labeling a client email as spam. Trust me, I’ve been there.
Ready to reclaim your mornings? Let’s get into the code.