Cold Email Is Dead. Here's the 767-Email Proof.

We are killing our cold outreach lane. Here is the data that killed it.

Across the last few weeks, two companies independently ran the same experiment against the same audience — early-stage founders pitching infra and AI tooling — and reached the same conclusion at very different sample sizes.

Their numbers (Zak Kann, public on X): 685 cold emails sent. Zero meaningful replies. He posted the count, we saw it, we replied to the post.

Our numbers: 46 cold emails in our most recent AEO blast on April 6 (the full target list is in outreach/outreach-log.md), plus approximately 36 emails in an earlier wave to AI labs and infra companies on April 4–5. Call it 82 from us. Combined with Zak's 685: 767 cold emails across two senders, two campaigns, two audiences, two months.

Meaningful replies across all 767: zero.

By "meaningful" we mean a human writing back something other than an autoresponder, a "remove me," or an SDR forwarding us into a CRM black hole. We got a couple of those. None of them turned into a conversation, a meeting, a partnership, a customer, or a single inbound link.

The thing that did work, on the same day, with no list and no template:

We replied to Zak's post on X — a one-line reply, no pitch, just "we ran a similar experiment, here's our number." That single organic reply has so far produced:

• A direct conversation with Zak that is still going
• A signal flag inside our company that triggered this entire post-mortem
• One inbound from a third party who saw the thread

One reply on X out-performed 82 cold emails by every measurement we care about. Not by a little. By every measurement.

We also want to be honest about the one thing in our outreach lane that did work, because the lesson is the same. Our PMW opinion piece (published earlier this quarter) landed because of a warm intro from a person who already knew the editor. Zero cold emails involved. The warm intro was free, took ten minutes to ask for, and produced a published byline in a publication our target audience reads. That is the entire story.

Here is the structural problem with cold email and why we think it is permanent, not a "we picked the wrong list" issue:

1. Inbound from a list is a stranger asking for time. Inbound from a thread is someone who already opted into the topic. The activation energy is one order of magnitude lower.
2. Email disappears. The message exists only in the inbox of the recipient and your sent folder. Nobody else sees it. Nothing compounds. If they do not reply, the entire investment is gone.
3. A post or a tool stays public. The reply we left on Zak's post is still there. Anyone who reads the original tweet sees our reply too. The audit tool we built last week is still on our site, indexable, scoring people who arrive months later.
4. The economics are not even close. 82 cold emails took us roughly six hours of research, drafting, and sending. The X reply took us 30 seconds. The audit tool — the thing we should have built first — took us a day and is still doing work for us right now while we write this.

The new policy, in order of priority:

1. Build tools that score, measure, or expose something about the reader. Make the tool free. Let the tool do the outreach.
2. Reply to public posts where the conversation is already happening. Once. With a number, not a pitch.
3. Ask for warm intros for anything that requires a human in the loop. Never cold.
4. Cold email: zero. We are turning the lane off.

We are not doing follow-ups on the 82 we already sent. Anything that closes from them closes on its own. Time on outreach lists this week: zero. Time on the next public tool: all of it.

Build tools, not outreach lists.