I Get 40 Cold Emails a Day — Here's How I Got It to Zero

I'm the kind of person who compulsively reads every email. Doesn't matter if it looks like spam - something in my brain registers unread counts as unfinished work. For most of my career this was a minor quirk. Then I started a company, and it became a problem.

The Day I Counted

About eight months ago I sat down on a Monday morning and actually counted. Not the emails in my inbox - the cold emails. I went through my Primary tab and flagged every message that was unsolicited outreach from someone I'd never heard of, trying to sell me something I hadn't asked about.

Forty-three. In a single day.

I'd known it was bad. I hadn't known it was that bad. The breakdown was roughly what you'd expect if you've ever run a company and been listed as a founder on LinkedIn or Crunchbase:

• SDR pitch emails: The classic format. "Hi [First Name], I noticed your team is growing and thought you might be interested in..." These came from real domains, had proper formatting, and referenced actual details about our company - funding round, headcount, tech stack. At least a dozen per day.
• Partnership proposals: "We work with companies like yours and would love to explore a mutually beneficial..." These are sales emails wearing a different hat. Same volume, same playbook.
• "I noticed you raised" emails: The moment our seed round was announced, something triggered in every database that cold email senders use. Apollo, ZoomInfo, Crunchbase enrichment feeds - something surfaced us to hundreds of BDR lists simultaneously. The first week after the announcement I received roughly 80 cold emails in five days.
• SEO/content/design agencies: A subcategory of their own. "I was looking at your website and noticed a few things..." Always vague, always a pitch.
• AI tool demos: Ironically, many of the most AI-generated emails I received were from companies selling AI tools. Meta.

The forty-three emails consumed somewhere between 45 minutes and an hour of my day. Not because I was reading them carefully - but because each one required a few seconds of evaluation, a decision (delete? report spam? block?), and the cognitive cost of context-switching back to whatever I was actually trying to do.

The Manual Methods I Tried

My first instinct was to fix this with Gmail's built-in tools. I am technically capable. Filters are not complicated. I spent an evening setting up rules.

Gmail filters caught some of it. I created filters for common phrases: "quick question", "15 minutes", "mutual benefit", "I came across your profile". This worked for maybe 20% of the volume. The other 80% used different phrasing - because modern cold email tools are specifically designed to avoid predictable trigger phrases, and because AI-generated content varies enough that no fixed phrase catches everything. I added more phrases. I played whack-a-mole for a week before admitting it wasn't going to scale.

Blocking individual senders was worse than useless. Each blocked sender was replaced by two more the next day. Senders frequently rotate domains - the same BDR reaching out from acmesolutions.io on Monday might use acme-growth.com on Wednesday. Blocking addresses one at a time is like bailing out a boat with a thimble.

Unsubscribing I tried exactly once and quickly stopped. On three separate occasions, clicking an unsubscribe link in a cold email resulted in a noticeable uptick in cold emails from different senders in the following week. I'm not certain these are directly connected - it's hard to prove causality - but the most likely explanation is that unsubscribe interactions confirm to data brokers that the email address is active, which increases its value on contact lists. I stopped unsubscribing entirely.

Email aliases were the most useful manual technique. I started using firstname+conference@domain.com for conference registrations, firstname+content@domain.com for content downloads, and so on. This let me identify which services were selling or leaking my email. Turns out: almost all of them. The alias trick is useful for limiting future exposure, but it doesn't help with the historical problem - my real email address was already in every database that mattered.

Understanding the Cold Email Machine

After a few weeks of failed manual fixes, I got curious about the supply side of this problem. What exactly was generating 40+ emails per day?

The answer turned out to be a mature, well-funded industry I'd been blissfully unaware of. The cold email industrial complex is enormous: Apollo has over 500,000 users. Instantly and Smartlead are growing at venture-backed rates. Lemlist, Reply.io, Saleshandy - the list of platforms specifically designed to flood professional inboxes goes on for pages. The economics are stark: a BDR team can purchase a $100/month subscription to one of these tools and reach 10,000 targeted prospects. The math strongly favors sending more.

What I hadn't fully understood was how personalized these tools had become. I used to be able to spot a cold email immediately - something in the generic phrasing, the obvious template structure, the slightly-off personalization. That was 2023. By 2025, these emails had crossed a threshold. I was reading them and genuinely wondering for a few seconds whether I actually knew this person.

The AI personalization pipeline works like this: the tool ingests your LinkedIn profile, your company's website, any recent press coverage, your Twitter/X activity if you're public. An LLM generates an opening line that sounds like the sender did actual research: "Congrats on the product launch last month - the approach to [specific feature] is interesting." The rest of the email is templated but the hook reads as genuine. I fell for it multiple times, opening emails and reading past the first paragraph before realizing it was outreach.

Once you understand the infrastructure behind cold outreach, the failure of manual filtering methods makes sense. You're not filtering emails from individual senders - you're attempting to manually defeat a purpose-built industrial system. It's not a fair fight.

The Tool That Actually Worked

I found Email Ferret through a thread on a founder Slack I'm in. Someone posted about it with the kind of understated enthusiasm that tends to indicate genuine product satisfaction ("it just... works"). I was skeptical enough to spend twenty minutes reading about how it detected cold outreach before signing up.

Setup took about five minutes. Authenticate with Gmail, configure the threshold (I kept the default, which flags emails scoring 4/10 or higher on the spam scale), and let it run. No rule configuration. No phrase lists. No maintenance.

The first week results stopped me cold: 38 of 40 daily cold emails caught and labeled. Automatically. Moved to a "Cold Outreach" label and archived, out of my Primary tab, without me lifting a finger.

More importantly: I did not miss a single legitimate email. This was my primary fear with any automated filtering system - that something real would get caught and I'd miss a customer email or an investor note. Over seven weeks of use, I've had zero false positives that I've been able to identify. The system is conservative enough that if there's any signal of a real relationship - previous contact, trusted domain, same company - the email comes through to Primary.

The two emails per day that did get through to my inbox were genuinely ambiguous cases: one was from a founder I'd met at a conference who was now at a different company and reaching out about a potential partnership (real relationship, legitimate outreach), and one was from a domain that turned out to be a legitimate vendor. Both correctly landed in Primary.

What Email Ferret is doing under the hood is evaluating a combination of 15+ signals - domain age, BDR phrase density, whether the sender has ever emailed anyone I know, the presence of cold outreach tool headers, a score for "sales intent" from an LLM check - and combining them into a composite score. No single signal catches everything. But the combination catches almost everything that individually looks clean.

My Inbox Now

Eight weeks after setting up Email Ferret, my Primary tab looks like what email was supposed to look like: messages from people I actually know and services I actually use.

The cold outreach still arrives - I'm not naive enough to think the sending has stopped. It just doesn't reach me anymore. The "Cold Outreach" label in Gmail has 200-300 new messages per week that I never see in my Primary tab. I check it occasionally when I'm curious, and the accuracy is remarkable - scrolling through a week of flagged emails, I've found maybe two or three that I'd consider genuinely borderline.

The time savings are real and substantial. My estimate is 30-40 minutes per day I'm no longer spending processing junk. That's roughly 150 hours per year - nearly four full weeks of working hours - that were previously being consumed by other people's sales pipelines and are now available for actual work.

More abstractly, something about the quality of my relationship with email has changed. I used to approach my inbox with low-grade dread - every check involved scanning for cold outreach, evaluating each message, and context-switching repeatedly. Now when I open Gmail, I mostly see things I actually want to see. That's a more significant change than the raw time numbers suggest.

What I Learned

A few things I'd tell myself eight months ago:

This problem is structural, not fixable through discipline. I spent months trying to manage cold email volume through better habits and manual filtering. The volume is engineered to be unmanageable by any reasonable individual effort. You need a tool that matches the sophistication of the sending infrastructure.

The problem will get worse, not better. AI writing tools are becoming cheaper and more capable. Cold email platforms are adding more AI features every quarter. The barrier to sending 500 personalized cold emails per day continues to fall. Manual defenses that barely work now will be completely overwhelmed within 18 months.

Unsubscribing makes it worse. This took me too long to accept. Don't engage with the unsubscribe mechanism in cold emails from senders you don't recognize. Delete, or let your filter handle it.

The economics favor proactive defense. The cold email industry is operating on the assumption that their cost of sending is far lower than your cost of receiving. At $100/month, a BDR team can spam 10,000 people. At $5/month, you can stop nearly all of it from reaching you. The defensive economics are better than the offensive economics. Use that advantage.

The guide on blocking cold emails in Gmail covers the setup process in detail if you want step-by-step instructions. The short version: stop trying to manually filter, add AI detection, and get back the 30+ minutes per day you're currently spending on other people's sales pipelines.