How to Warm Up Email Domains and Mailboxes

Gmail, Outlook and Yahoo score senders on observed behaviour: how much you send, how consistently, how recipients react, and how many messages bounce or get marked as spam. A new domain has none of that history, so filters apply the harshest defaults available. Warmup manufactures the missing history — a small, steadily increasing stream of messages that get opened, replied to and dragged out of spam by a network of cooperating mailboxes.

Two things follow from that. First, warmup is behavioural, not technical: no DNS record makes a cold domain trusted. Second, it is never finished. Reputation decays when sending stops and drops sharply after a bad campaign, which is why mature outbound teams keep a background warm running on every mailbox permanently.

Never run cold outbound from the domain your customers, invoices and support email depend on. Buy lookalike sending domains — the `.co`, `.io`, `get-`, or `try-` variants of your brand — and treat each as disposable infrastructure. If one burns, you retire it without touching the business.

Each sending domain needs full authentication before warmup begins. SPF lists who may send for the domain, DKIM cryptographically signs each message, and DMARC tells receivers what to do when either fails. Google and Yahoo now require all three from bulk senders, so an unauthenticated domain is not merely disadvantaged — it is filtered.

The exact numbers matter less than the shape: start tiny, increase gradually, never spike. Below is the ramp we recommend per mailbox. Campaign sends are additive to warmup traffic, so count both against the daily figure.

A sequencer's built-in warmup is fine for one or two mailboxes. Once you run several — and any serious outbound programme does — you want warming that is independent of the sending tool, so switching sequencers never resets your reputation work.

Warmforge is the dedicated service we recommend: continuous warming across every mailbox, automatic ramping, inbox-placement and reputation monitoring, and alerts when a mailbox starts landing in spam. Because it sits outside the sequencer, the same warm pool covers mailboxes whether you send through Lemlist or Apollo.

Whatever you use, connect every mailbox — including the quiet ones. Reputation is scored per mailbox, and an unwarmed mailbox added to a rotation drags the whole campaign's delivery down.

Open rate is no longer trustworthy: privacy proxies pre-fetch images and inflate it. Watch the metrics providers actually act on. Google's Postmaster Tools reports domain reputation and complaint rate directly from Gmail, which is the single most useful free signal you have.

If placement collapses, stop campaigns on that domain immediately — continued sending deepens the hole. Verify the list you were sending to, remove every stale or risky address, and check that authentication has not broken. Then return the mailbox to warmup-only traffic for several weeks before attempting any campaign volume again.

Be honest about the economics. Recovery takes weeks with no guarantee, and a fresh sending domain costs roughly the price of a coffee. If a domain has been badly burned, retire it, spin up a replacement, and this time warm it properly before the first campaign.

Warmup and list hygiene are the same project. Verify addresses before every send at volume — contacts change jobs constantly and a 'verified' list ages within weeks. Maintain real suppression: unsubscribes, current customers, open opportunities, competitors and anyone who has asked not to be contacted.

One bad list can erase a month of warming in a single morning. The cheapest deliverability spend in any stack is verification, followed by warmup, followed a long way back by everything else.