Email warmup explained: what it does, and why you can't skip it

A new mailbox that sends fifty cold emails on its first morning looks exactly like a spammer. Nothing about the messages has to be bad. The problem is that a mail provider has no history with the sender, no evidence that anyone wants to hear from it, and a sudden burst of unfamiliar mail is one of the cheapest patterns for a filter to act on. Email warmup exists to give that mailbox a history before it has to earn one from strangers.

Most people explain warmup as “tricking spam filters,” and that framing causes two mistakes. Teams either treat warmup as a magic switch they flip once and forget, or they dismiss it as a fake signal that providers will eventually see through. Both views miss what it does. Warmup is a way to create a believable sending pattern: steady volume, recipients who engage, and mail that doesn’t bounce or get reported. Whether that pattern is enough depends on everything else you do.

This article explains what warmup actually signals, how the different kinds of warmup networks differ, how to design a ramp you can defend, and what warmup cannot fix. If you’re setting up a new sending domain or mailbox, this is the part of the process that most often goes wrong, usually because someone decided it was optional.

What email warmup actually signals

Mail providers don’t know whether you’re a person, and they don’t need to. What they can observe is how recipients behave when mail from your address arrives. Did the recipient open it, reply, star it, move it out of spam, or mark it as spam? Did the address bounce? Did the volume grow at a pace that matched the sender’s history? Over time those observations form a reputation for the sending domain and, to a lesser degree, the sending IP address.

A brand-new domain has no observations at all. Providers have to fall back on priors, and their priors for unknown senders are cautious. The first few hundred messages are the most scrutinized, because the filter is trying to decide what kind of sender it’s dealing with.

Warmup fills that gap by generating the kind of observations a legitimate sender would collect. A warmup network sends messages from your mailbox to other mailboxes, and those mailboxes respond: they open, reply, mark as important, or move the message to the inbox from spam. The sender’s history begins to look like a sender whose mail people want.

The honest limitation is in that last sentence. Warmup emulates engagement. It can’t manufacture the behavior of real recipients who have a reason to care about your message. A mailbox that warms up beautifully and then sends its first real campaign to a list of people who never asked to hear from it will still struggle, because the signal that matters most, real recipients engaging with real offers, is absent. Warmup is a bridge. It gets you to the point where real engagement can take over, and it does not replace that engagement.

The three kinds of warmup

There are three common approaches, and they differ in cost, control, and how the signal looks to a filter.

In practice, most teams use a peer network for volume and add manual warmup for the first few weeks, when a real human reply is worth more than dozens of automated ones. Seed accounts are useful mostly for diagnosis. If you want to know whether a new mailbox lands in the inbox or in spam at Gmail or Outlook, seed accounts at those providers will tell you, and they’ll tell you faster than waiting for campaign results.

Be skeptical of any warmup that promises to make your mail look “perfect” or that claims a score of 100. A score is the tool’s internal measure of its own activity. It says nothing guaranteed about how a provider classifies your mail. Warmup networks that fake opens from unrelated accounts or that rotate through patterns no human would produce are a liability, because when the pattern is discovered, the mailbox loses the history you were paying to build.

A ramp you can defend

The most useful thing email warmup teaches is pacing. A mailbox’s daily volume should rise gradually, and each step should be held until the numbers stay healthy. Here is an illustrative schedule for a single mailbox on a domain that has been authenticated and registered for a few weeks:

Week one starts at 5 to 10 messages a day, mostly warmup traffic with a handful of manual replies. Week two moves to roughly 15 to 25. Week three reaches 30 to 40, week four 50 to 60, and week five can approach 70, at which point cold sends can start to carry real volume if the numbers have held. These figures are a shape rather than a standard. A mailbox on a fresh domain with a clean list may need longer. A mailbox on an established domain with a history of good engagement may move faster.

The rule I use is simple: I rarely let a mailbox’s daily volume grow by more than about 25 percent from one day to the next, and I never double it in a day. Doubling is the pattern that looks like a script being turned on. A slower curve also gives you time to catch problems while they are still small. If bounce rate creeps up or a complaint shows up in a segment, you want to notice on day nine, not on day twenty-one when the mailbox has already started to slide.

Hold steady at each step until three things are true at once: hard bounces are low, complaints are close to zero, and the warmup or manual replies are arriving as expected. Stepping up because the calendar says it’s time is how teams end up in spam during week four.

What warmup cannot fix

The most expensive misconception about email warmup is that it repairs whatever is wrong with the sending setup. It doesn’t. Warmup works on reputation, and reputation is a running record of what happens when mail is sent. If the mail itself has problems, warmup gives those problems more time to accumulate.

A list with a high share of invalid addresses will bounce regardless of how well the mailbox is warmed. Each bounce is a negative signal, and a warmed mailbox that suddenly starts bouncing is more alarming to a filter than a cold one that bounces from the start. Verify the list before warmup, not during it.

A domain without SPF, DKIM and DMARC is also beyond warmup’s reach. Authentication is the baseline a provider checks before it looks at history at all. If the records are missing or misaligned, the warmed mailbox still has an identity problem. We wrote about the three records in SPF, DKIM and DMARC explained, and the short version is that you should confirm they pass before the first warmup message goes out.

A domain or IP that is already on a blocklist will stay penalized while warmup runs. Check Spamhaus and the major provider lookups before you start. If the domain has an unknown history, because it was bought secondhand or used by a previous company, assume its reputation is an open question until the data says otherwise.

Finally, warmup does not make bad content acceptable. Mail that is image-heavy, link-stuffed or obviously templated can still be filtered after warmup. The warmup history gives you the benefit of the doubt on sender identity. It doesn’t rewrite the message.

What to measure while you warm up

Warmup generates its own noise, so the numbers you care about are the ones that come from the outside world. Watch hard bounce rate, which should stay well under 2 percent. Watch spam complaints, which Google describes as a rate to keep below 0.1 percent and never let approach 0.3 percent. Those two thresholds come from Google’s bulk-sender guidance, and they matter whether or not you’re a bulk sender.

Google Postmaster Tools is the most useful free instrument here, and it’s worth setting up before your first send. It reports domain reputation as High, Medium, Low or Bad, along with spam rate and authentication pass rates. One caveat: Postmaster only shows data once a domain sends enough mail to Gmail, so a new domain will show no data for a while. That is not a sign of a problem. It means the sample is too small to report.

Seed testing adds a second view. Send a test message to a few accounts you control at each major provider and check where it lands. Do this at the start of each ramp step rather than only at the end. Placement tests are noisy, so treat them as a direction, not a verdict. If the same mailbox lands in spam for three tests in a row at one provider, act on that.

Warmup and real engagement

The most important question in warmup is what happens after it ends. Some teams treat warmup as a phase: run it for a month, then switch it off and send as if the reputation is now permanent. Others never switch it off. Both approaches miss the same point. Reputation is kept alive by real recipients, and the share of real engagement should grow as warmup traffic shrinks.

My position is that email warmup should taper, not end abruptly. Many teams keep it running at a low level after cold sends start, which keeps the mailbox from going silent in a way that is hard to explain to a filter. I’d rather keep a light warmup running than find out three weeks later that a mailbox which used to be engaged has gone quiet and started landing in spam. That is a judgment call, and reasonable people set it differently, but the failure mode of switching off everything at once is real.

Real replies are worth more than warmup replies, which is why manual warmup in the first few weeks is worth the effort. A colleague who replies to a test message from a new address is telling a filter something no network can fully simulate.

Edge cases that change the plan

A few situations need a different approach. Reactivated domains, meaning domains that used to send mail and then went quiet, carry their old reputation, good or bad. Look at the history in Postmaster and the blocklists first, then decide whether to warm or to retire the domain. A domain with a poor history should not be warmed up to hide its past.

Shared sending infrastructure changes the picture too. If you send through an ESP’s shared IP pool, your IP reputation is partly determined by other customers. In that case the domain’s own reputation and authentication matter more, because they’re the parts you control.

Mailboxes that send from a fresh domain registered only days ago deserve extra caution. Filters are skeptical of very new domains, and a domain that has been registered for a few weeks with a plain website and no sending history will warm more slowly than an older one. Give the domain time before you start the ramp, and do not treat registration age as a loophole.

If you run several mailboxes on one domain, ramp them together but not identically. Filters see the domain as the unit, so a sudden spike across all mailboxes at once looks more like a campaign than a set of individual senders.

Wrap-up

Warmup is a sending history you build on purpose. It works because providers judge senders by how recipients behave, and it gives a new mailbox a believable pattern to start from. It works only if the list is clean, the records pass, the domain has no blocklist baggage, and the ramp is slow enough to read. Ramp by roughly a quarter per day, hold until the numbers are flat, measure what providers actually report, and taper warmup as real engagement takes over.

If you’d rather not assemble that stack yourself, OutboundLeadPro runs cold outreach from your browser with unlimited mailboxes and warmup included, plus DNS auto-fix for SPF, DKIM and DMARC, so the setup work is handled before your first send. You can see how it fits together on the features page. For the wider picture of what decides inbox placement, our deliverability checklist is the next read.

Keep reading

Playbook

Cold Email A/B Testing: Why Most Tests Prove Nothing

Cold email A/B testing is often too small to mean anything. Here is the sample-size math at real reply rates, and how to read the results.

Daniel Khoury September 19, 20263 min read

Playbook

How to avoid spam folder placement: 12 causes and fixes

Want to avoid spam folder placement? These 12 causes, from authentication to list quality to complaints, each come with a concrete fix you can check.

Daniel Khoury September 24, 20263 min read

Playbook

How to build a lead list that stays clean after the first send

Build a lead list around a real ICP, with stage gates that reject bad rows before they cause bounces and complaints. Includes a worked example.

Daniel Khoury October 7, 20263 min read

Explainer

SPF, DKIM and DMARC explained for cold email

The three email authentication records decide whether you reach the inbox — and most cold outreach fails because one is missing or wrong. Here's each one in plain English.

The OutboundLeadPro Team October 11, 20262 min read