The concrete leveling proof sequence

Three emails, the DNS records that have to exist first, and a 14-day warmup. Written 2026-07-31.

Nothing here can send yet, and nothing here has been sent. stpierre.ai has no DKIM and no DMARC — only SPF is published. Cold email off the main domain would burn it. The sequence goes live about three weeks after Akash buys a separate sending domain and starts the warmup below. Buying the domain is his call and his money.
3
emails
6
DNS records
14
warmup days
~3 wks
until first send

Rules the copy follows

The three emails

Send days 1, 4, and 8. Same thread if your tool supports it. If they reply anything at all, stop the sequence and answer as a human.

Email 1 — the Dallas tape
Subject lines (pick one, test the others later)
Body — plain text, no images, no HTML, no tracking pixel
Hey {{first_name}}, I run the Facebook ads for a concrete leveling contractor in Dallas. Here is what eight days looked like. 33 homeowner conversations. 16 of them qualified. He spent $406.62 total. That is about $25 a qualified lead. SCREENSHOT 1 - the transferred leads landing on his phone SCREENSHOT 2 - the live dashboard, Jul 21 to Jul 29 I am not going to dress it up. That is one city, eight days, and one meeting booked so far. But I looked, and not one marketing company in leveling has ever published a real client number. So I am publishing mine. Want the dashboard link? Akash St. Pierre {{phone}} Reply STOP and I will never email you again. UNSUBSCRIBE LINK
Email 2 — the free chatbot
Subject lines (pick one, test the others later)
Body — plain text, no images, no HTML, no tracking pixel
Hey {{first_name}}, Different email. This one costs you nothing. I build a chatbot that sits on a concrete leveling website and books the homeowner while you are on a jobsite. It asks what is sinking, gets the address and the phone number, and pushes the lead wherever you already keep them. I will put one on {{company}}'s actual site and send you the link. Not a slide deck. Your real site. Takes me about a minute. It is yours whether or not we ever talk again. DEMO LINK - stpierre-demo-bot.pages.dev/{{slug}} Want it? Akash St. Pierre {{phone}} Reply STOP and I will never email you again. UNSUBSCRIBE LINK
Email 3 — the guarantee
Subject lines (pick one, test the others later)
Body — plain text, no images, no HTML, no tracking pixel
Hey {{first_name}}, Last one from me. Here is the whole deal in plain English. You put down $2,000. I build the system, wire it into your CRM, and run the ads. If I do not put $20,000 of sold concrete leveling work on your books in 30 days, you get the $2,000 back. All of it. After that I take 10% of what actually closes from the work I brought in. Nothing closes, I do not get paid. I am not a lead vendor. A lead vendor sells the same homeowner to you and three competitors. PROOF CARD - Dallas: 33 conversations, 16 qualified, about $25 a qualified lead Worth 30 minutes? Akash St. Pierre {{phone}} Reply STOP and I will never email you again. UNSUBSCRIBE LINK
Placeholder slots to fill before the first send. Screenshot 1 and 2 are real captures from the Dallas account — take them fresh the day you send, not from an old file. The demo link is minted per prospect off stpierre-demo-bot.pages.dev. The proof card in email 3 is the same Dallas numbers, styled. If a slot is still empty on send day, cut that line rather than describe a screenshot you did not attach.
One thing to reconcile before this ships. The outbound playbook says zero upfront and Akash fronts the ad spend. Email 3 asks for a $2,000 deposit that is fully refundable. Both can be true — the deposit is refundable, so the risk is still zero — but the phone script and the email have to say the same thing. Pick one wording and use it everywhere.

DNS records for the sending domain

Buy a look-alike, never the main domain. Something like getstpierre.com or stpierremarketing.com. Everything below is templated — swap the bracketed parts once the domain and the sending tool are chosen.

WhatTypeHostValueWhy it matters
SPFTXT@ (root of sending domain)v=spf1 include:<ESP-SPF-INCLUDE> ~allOne SPF record only. Never two. Swap the include for whatever sending tool you pick. Keep ~all (softfail) through warmup, tighten to -all after day 30.
DKIMTXT<selector>._domainkeyv=DKIM1; k=rsa; p=<2048-BIT-PUBLIC-KEY>2048-bit, not 1024. The sending tool generates the key and gives you the selector. Some tools hand you a CNAME instead of a TXT — use whichever they give you, do not publish both.
DMARCTXT_dmarcv=DMARC1; p=none; rua=mailto:dmarc@<SENDING-DOMAIN>; fo=1; adkim=r; aspf=r; pct=100Start at p=none so nothing gets blocked while you watch the reports. Move to p=quarantine around day 21 once SPF and DKIM both pass clean.
Return-pathCNAME<bounce-subdomain> (e.g. bounce)<ESP-BOUNCE-TARGET>This is what makes SPF line up with the From address. Skipping it is the most common reason a technically correct setup still lands in spam.
MXMX@ (root of sending domain)<MAIL-HOST> (priority 1)The sending domain has to be able to receive mail, or replies and bounces vanish and your reputation reads as a one-way spam cannon.
A recordA@ (root of sending domain)<WEB-HOST-IP>Point it at a real one-page site. A sending domain that resolves to nothing looks disposable to every filter.
Check your work at mxtoolbox.com and dmarcian.com before day 1. All three of SPF, DKIM, and DMARC have to read pass on a test send to a Gmail address, or the warmup is just teaching Google to distrust you.

14-day warmup

One inbox on the new domain. Sends per day, and what that day is for.

DaySendsToWhat you are doing
15Seed inboxes onlyGmail, Outlook, Yahoo accounts you own. Open, reply, mark not-spam, drag out of Promotions.
25Seed inboxesSame. Reply to every one from the seed side.
310Seed inboxesAdd a second Gmail seed. Start a real thread, two replies deep.
410Seed + 2 friendliesFirst outside humans. People who will actually reply.
515Seed + friendliesCheck the DMARC reports. SPF and DKIM should both read pass.
615FriendliesNo weekend sends after this point unless replies are strong.
720FriendliesWeek-1 review: any bounce over 3% means stop and fix the list.
825First cold batch20 warm + 5 cold. Cold rows must be verified addresses only.
930Cold batch25 cold. Same subject family all week, so you can read the numbers.
1035Cold batchWatch reply rate, not open rate. Opens are unreliable now.
1140Cold batchAny spam complaint at all = pause the day, check the source of those rows.
1250Cold batchMove DMARC toward p=quarantine if SPF and DKIM have passed clean all week.
1360Cold batchAdd the second subject line. One variable at a time.
1475Cold batchWeek-2 review. If bounce is under 2% and complaints are zero, keep climbing ~20% a day to a 150/day ceiling per inbox.
Auto-stop rules. Bounce rate over 3% — stop sending, re-verify the list. Any spam complaint — pause the day and check where those rows came from. Zero replies after 50 cold sends — change the subject line or the list, not the volume.
Built 2026-07-31 for the B2B outbound launch lane. Sources: Outbound Playbook v4 (Jul-29), the CL market map (Jul-28), and the live Dallas numbers off Meta account act_3289323034532141 cross-checked against sp_client_leads. Zero emails were sent. No domain was purchased.