VIRAL TILT · OUR OWN STORY · 2024–2026
THE QUESTION
THE ANSWER
177,693COLD EMAILS SENT FROM OUR OWN INBOXES, 2024 TO 2026 · THE SCALE, NOT THE RESULT
> WHAT WE WALKED INTO
This file carries our own name because the partner is us. Viral Tilt’s outbound has been written, warmed, and sent in-house since 2024: 177,693 sends through more than 60 inboxes over that time, on our own domains and our own reputation. The send count is the number every agency puts on its site, and it is the least useful one on this page, because sends are what you paid for and replies are what you got. So here are the replies, from the program we run today, at the low end of the published benchmarks and stated that way.
> THE MOMENT · THE MACHINE, AS LOGGED
> STEP 01
Our name in the from line, our domains carrying the reputation, 60-plus inboxes over three years, rotated as they wear out, because inboxes and domains are consumables and the senders who treat them as assets are the ones that get filtered.
> STEP 02
Sequencers count every autoresponder as a reply, so we read human replies by hand. In the 2026 program, about 23,000 sends produced a human reply rate of 0.26 to 0.28 percent. Published averages run 3 to 4 percent. Ours is a cold-scraped list and a new offer, and the number says that louder than any excuse we could add.
> STEP 03
Every one of the 15 interested leads in ten weeks answered an offer-first ask: we build the thing, then offer to send it. The plain “worth a chat” versions produced zero. That split is the most useful thing we learned in three years of sending, and it cost 23,000 emails to learn.
> STEP 04
A free MX and DNS pre-screen took bounces from 11.4 percent to 1.8 percent with no paid verification. Deliverability first, list second, copy third, which is the order the numbers forced on us.
> THE NUMBERS, DRAWN FROM THIS FILE ONLY
SEND AND INBOX COUNTS FROM OUR OWN SENDING RECORDS · PROGRAM FIGURES FROM OUR 2026 CAMPAIGN LOGS
OUR OWN NUMBERS, INCLUDING THE UNFLATTERING ONE.
> WHAT CHANGED · FLIP IT YOURSELF