Cold email fails in the plumbing long before the copy matters. The pattern is always the same: someone sends 500 emails from their main domain on day one, lands in spam, poisons the domain their invoices come from, and concludes “cold email doesn’t work.” The infrastructure below is what actually decides deliverability — and none of it is secret, just unglamorous.
Rule zero: never send cold from your real domain
Your primary domain carries your business mail — invoices, client threads, password resets. Cold outreach carries spam-complaint risk by nature. Separate them absolutely: buy adjacent domains for outreach so the worst case burns an $11 asset instead of your company’s ability to email its own customers.
The authentication trio
Every sending domain needs three DNS records before its first email: SPF (which servers may send for you), DKIM (a signature proving mail wasn’t forged), and DMARC (the policy tying them together). Missing records don’t just hurt — mail providers increasingly refuse bulk senders without them outright. Verify with our deliverability checker in seconds.
Warmup is not optional
A brand-new inbox that sends 50 emails on day one looks exactly like a spammer, because that’s what spammers do. Warmup builds a sending history first: tiny volumes, gradually increased, with realistic engagement. Our planning rule: 14 days minimum before real campaigns, 21–30 days when the list matters — that’s our recommendation from operating experience, and the reasoning is simply that reputation systems reward age and consistency, and nothing else substitutes for them.
The capacity math
Deliverability is a per-inbox, per-domain game, so volume comes from multiplication, not aggression. Our conservative planning defaults: about 30 cold sends per inbox per day, and about 3 inboxes per domain. Need 300 sends a day? That’s ten inboxes across four domains — not one inbox sending 300. The infrastructure planner does this arithmetic with editable assumptions, because these are defaults, not laws.
Blacklists: monitoring, not mythology
Public DNS blacklists (Spamhaus, SURBL, URIBL and dozens of smaller ones) are queried by real mail filters in real time. Young lookalike domains — the kind cold outreach uses — trip them more easily than aged ones. Getting listed isn’t a scarlet letter; it’s a signal to act: pause that domain, keep it warming, and use each list’s free delisting process. What’s inexcusable is not knowing — check monthly, it takes seconds. (Full detail: how blacklists actually work.)
List hygiene is deliverability
Every hard bounce tells providers you’re careless with data. Verify addresses before sending — verification services are cheap compared to a burned domain — and treat a campaign bouncing above roughly 5% as a stop-everything signal to fix the list, not push through.
The compliance floor
U.S. commercial email is governed by CAN-SPAM: accurate sender information, no deceptive subject lines, a working opt-out honored promptly, and a physical mailing address in the message. That’s a description of the law, not legal advice — rules differ sharply outside the U.S. (Canada and the EU are far stricter), so selling internationally is a talk-to-a-lawyer situation, genuinely.