Email Warmup
Warmup sends a slowly rising volume from a new address to mailboxes that reliably open and reply, building a sending history before you send anything that matters.
How long it takes: 21 days on a slow schedule, or 40 to 42 days on a normal one.
Does it work? Across 18 of our warmed mailboxes, 417 of 434 warmup emails landed in the inbox and 8 went to spam. What it cannot fix: broken authentication. Six of 19 sending domains we hold health data for had an invalid SPF or DMARC record, and no amount of warmup helps those.
Email warmup is one of those practices that everybody in cold outreach agrees is mandatory and almost nobody publishes numbers for. This page reports ours. We run warmup on customer sending mailboxes as part of the product, so the placement and configuration figures below come from our own production records rather than a vendor's marketing page.
Two things worth saying up front. The samples here are small, tens of mailboxes rather than thousands, and we say so at each number. And this is not a tool roundup: if you want products compared on published rates, that is best email warmup tools.
What warmup actually does
A mailbox provider decides where your mail lands using, among other things, the history of the address sending it. A brand new address has no history. Its first significant action being a burst of outbound mail to strangers is behaviourally identical to a spammer buying a fresh domain, and it gets treated accordingly.
Warmup manufactures the missing history. The address sends a small and slowly increasing number of messages into a pool of cooperating mailboxes, which open them, reply to some, and pull anything that lands in spam back into the inbox. Those are the positive engagement signals a provider uses to decide you are a real correspondent rather than a list.
The important part is the shape of the curve, not the volume. What you are demonstrating is that the address behaves like a person's mailbox: a bit of mail, consistently, with replies coming back.
What our warmup data looks like
Across 18 warmed mailboxes with sends in a rolling 7-day window:
| Placement | Emails | Share |
|---|---|---|
| Inbox | 417 | 96.1% |
| Spam | 8 | 1.8% |
| Other (promotions, updates) | 9 | 2.1% |
| Total sent | 434 | 100% |
The distribution matters more than the average. Fifteen of the eighteen mailboxes recorded no spam placement at all. All eight spam landings came from three mailboxes, and one of those accounted for five of them. So the honest reading is not "warmup gets you to 96%". It is that warmup usually gets you to essentially 100%, and when it does not, the problem is concentrated in one mailbox with something specifically wrong with it rather than spread thinly across all of them.
That is a useful diagnostic. If your warmup dashboard shows spam placement scattered evenly across every mailbox, that is a domain-level or list-level problem. If it is concentrated in one, that mailbox has its own issue.
How long warmup takes
Three to six weeks, and the schedule is a deliberate choice rather than a guess. Across the mailboxes we run:
| Schedule | Duration | Mailboxes |
|---|---|---|
| Slow | 21 days | 5 |
| Normal | 40 to 42 days | 14 |
Anything advertising a warmed inbox in a few days is not warming it. The signal being constructed is a history across time, and time is the one input that cannot be compressed. A provider that has watched an address behave reasonably for five weeks has more evidence than one that watched it for five days, and there is no shortcut around that.
The practical consequence is a planning one: warmup has to start before you need the mailbox, not when you are ready to send. Six weeks ahead of a campaign is a comfortable margin.
The thing warmup cannot fix
Warmup builds reputation for an address that is otherwise correctly set up. It does nothing whatsoever for a domain whose authentication is broken, and broken authentication is far more common than people expect.
Of 19 sending domains where we hold health-check data:
| SPF / DMARC | Domains |
|---|---|
| Both valid | 13 |
| SPF invalid, DMARC valid | 2 |
| SPF valid, DMARC invalid | 2 |
| Both invalid | 2 |
Six of nineteen, near enough a third, had a broken SPF or DMARC record. Those domains can warm up for six weeks and still land in spam, because the receiving server cannot verify that the mail is authorised in the first place. Warmup is downstream of authentication, and running it on an unauthenticated domain is spending weeks to fix the wrong thing.
Check the records before you start the clock. It is a fifteen-minute job that determines whether the following six weeks are worth anything.
The other thing it cannot fix: your list
Reputation damage comes from two directions and warmup only addresses one. The other is bounces, and those come from the list rather than the mailbox.
Of 17,507 contact addresses we verified before sending, 17.1% came back invalid. Sending to an unverified list at that rate produces a bounce rate no warmed mailbox survives. Our own bounce rate is 2.8% across 11,477 sends, and getting there was about verification rather than warmup. The full breakdown is in cold email bounce rate.
Warmup and list hygiene solve different halves of the same outcome. Doing one and not the other gets you a well-regarded mailbox sending to dead addresses, or a clean list sent from an address nobody trusts.
When you can skip it
Rarely, but the case exists. A mailbox that already sends and receives normal business mail every day is being warmed continuously by its own use. Adding an artificial warmup pool to it adds little.
The situations where it is genuinely required are a new domain, a new mailbox on an existing domain, a mailbox that has been dormant for months, or one recovering from a reputation problem. Note the second of those: the address matters more than the domain. A fresh mailbox on a ten-year-old domain still has no sending history of its own. The established domain shortens the ramp rather than removing the need for it.
Where warmup sits in the wider picture
Warmup is one input into deliverability, not a synonym for it. The full stack is authentication, then a verified list, then warmup, then whether the message itself reads like something a person would send. Our broader treatment is in cold email deliverability, and if you want to check whether a specific address is worth sending to at all, the email verifier does that without signup.
Frequently asked questions
What is email warmup?
Sending a slowly increasing volume from a new address to mailboxes that reliably open, reply and rescue it from spam, so the address has a sending history before it sends anything that matters.
How long does email warmup take?
Three to six weeks. Our mailboxes run either 21 days on a slow schedule or 40 to 42 days on a normal one. Anything promising a few days is not warming anything.
Does email warmup actually work?
On our numbers, yes, with a small sample. 417 of 434 warmup emails across 18 mailboxes landed in the inbox, and 15 of those 18 mailboxes saw no spam placement at all.
What can email warmup not fix?
Broken authentication and a bad list. Six of 19 of our sending domains had invalid SPF or DMARC, and 17.1% of the addresses we verify are dead on arrival. Neither is a reputation problem.
Do you still need warmup on an established domain?
Usually yes, for a new mailbox on it. The address builds its own history. An old domain shortens the ramp rather than removing it.
Method
Placement figures cover 18 warmed mailboxes with sends recorded in a rolling 7-day window to 24 August 2026, 434 messages in total, from our production warmup records. Schedule figures cover 19 mailboxes with a configured ramp. Authentication figures cover 19 sending domains with a completed health check. These are small samples drawn from our own customer base and should be read as directional rather than as industry benchmarks; we are publishing them because almost nobody publishes any. Verification and bounce figures are from the separate dataset described in cold email bounce rate.