Private early access
MailRecon sends several message angles at once, learns which ones earn replies, and shifts volume to the winners. It verifies, paces and pauses to protect your domains. You approve the replies.
Used only to reply about access. Never shared.
The problem
You write a sequence, send it to a list and hope. When replies are thin you can’t tell whether it was the subject line, the angle, the follow-up or the list — so you can’t do more of what works.
And leads are finite. Every send that teaches you nothing is one you don’t get back.
MailRecon turns it into a running experiment, with a scoreboard you can trust.
How it works
Set the goal and approve the copy once. The loop runs from there.
Import a CSV, or have it pull fresh leads by country, city and vertical.
Every address is checked before the first email. Risky and catch-all results are held back, not mailed.
From your own mailboxes, in each lead’s timezone, ramped up gradually with daily caps and jittered timing.
Replies — and silence — score each angle. Volume shifts to what’s working; clear losers retire.
Replies are classified. Positives and questions get an AI draft for one-click approval, or auto-send if you’d rather. Out-of-office pauses a sequence instead of ending it.
Positive replies fire a webhook to your CRM — Kommo, n8n, Zapier, or anything that takes a POST.
What’s inside
Run several angles per campaign. A bandit allocates the sends, and once a loser retires, Claude breeds a fresh challenger from your best performer.
Describe your goal and offer; Claude drafts angles in your prospect’s language. Spintax and variable fallbacks keep every send unique — and never reading “Hi ,”.
Positive, negative, question, out-of-office, bounce, unsubscribe — classified automatically, with drafts waiting in a single inbox.
Describe who you want and it scrapes and verifies leads on demand, then queues them into the campaign you point it at.
Real-time webhooks push the whole record — email, phone, company and the reply itself — to wherever you keep deals.
One campaign per market, with copy and replies written natively rather than translated.
Guardrails
Most damage in cold email is done by a tool doing exactly what it was told. These are the checks that stand between a bad list and your domain.
Risky and catch-all addresses are held, not sent. You choose how strict to be.
New inboxes climb gradually. Daily caps, weekday-only sending, per-domain limits and send windows in the lead’s timezone apply automatically.
A bounce spike pauses the inbox. A high bounce rate across a campaign pauses the campaign. Either way, you get an email.
SPF, DKIM and DMARC for every sending domain, plus blacklist lookups against SURBL and Spamhaus.
An RFC 8058 header, an optional footer in your own words, and a suppression list every send is checked against.
Rates are per person reached, count genuine replies only — never bounces or auto-responders — and read “too few to read” until the sample means something.
Straight talk
You connect the mailboxes you already own, and your sender reputation stays yours.
That comes from your domains, your list and your offer. MailRecon makes the mechanical parts hard to get wrong.
Three replies is three replies, and it will tell you so.
Request access
Tell us a little about what you send and to whom — it helps us decide who to onboard first. Only your email is required.
Used only to reply about access. Never shared. No newsletter — a cold-email tool that spammed you would be a poor advert.