Email warmup: why it is not enough, and what to measure instead
Contents
How does email warmup work?Why does a good warmup score not guarantee the inbox?Can warmup harm your mailboxes?What should you measure instead of warmup?How do you decide a mailbox can send?Should you stop warmup altogether?In short
- Email warmup makes mailboxes exchange messages inside the tool's own network. It measures that network, not Gmail or Outlook.
- On our side, a warmup showed 100% inbox for weeks while every one of our tests at Google landed in spam.
- A warmup score of 100 can sit next to real sends that are all refused by the recipient's server.
- Warmup's automated traffic can get your mailboxes suspended by your host. Without it, our new mailboxes reached the inbox: 148 tests, 0 in spam.
- What really protects deliverability: measuring every week where your emails land, and sending only on what you observed.
Every new mailbox meant for prospecting comes with the same advice: "warm it up first". Email warmup promises to build the address's reputation before the first real send. We used it on our own mailboxes, measured it, then turned it off. Here is what we learnt, and what we do instead.
Every day Romeo spots the companies that need you and writes the first message. Nothing leaves without your approval.
Join the listHow does email warmup work?
A warmup tool plugs your mailbox into a network of thousands of other mailboxes. Every day they send each other messages, open them, reply, and pull them out of spam when they land there. The volume rises a little each day.
The idea is that mailbox providers see an active, opened, well-liked address. The limit fits in one sentence: that traffic stays inside the tool's network. It says almost nothing about what Gmail will do with a message sent to a prospect who does not know you.
Why does a good warmup score not guarantee the inbox?
Because it does not measure what a real send measures. We saw it on our first set of mailboxes, hosted on our own server.
- The warmup tool showed 100% inbox, week after week.
- At the same time, every placement test at Google, from September to early October, put the messages in spam: 13 mailboxes out of 13.
- Authentication (SPF, DKIM, DMARC), blacklists and reverse DNS were all clean.
What was left was the reputation of the server's IP address at Google. Warmup does not build it: Gmail judges an IP on the traffic it receives itself, not on exchanges between mailboxes of one network.
On some mailboxes, the warmup score stayed at 100 while every real send was refused by the recipient's server with a 554 error. The dashboard was green; the mailbox was sending nothing.
| What warmup measures | What a placement test measures | |
|---|---|---|
| Recipients | Mailboxes in the tool's network | Seed mailboxes at Gmail, Outlook and others |
| What is judged | Activity between friendly mailboxes | The real outcome: inbox or spam |
| IP reputation at Gmail | Not measured | Visible directly |
| Refused sends | Can go unnoticed | Show up as a failure |
| What we use it for | A volume indicator | The decision to send or not |
Can warmup harm your mailboxes?
Yes. Warmup traffic is automated, regular and voluminous, and some email hosts spot it. Our host suspended sending on several of our mailboxes because of that traffic.
So we turned warmup off on those mailboxes. The result speaks for itself: on a new set of domains, placement tests sent 148 seed messages, 0 in spam, at Google and Outlook alike. Those mailboxes send today without warmup.
What should you measure instead of warmup?
What matters is where your emails land, at which provider, this week. We check every sending domain weekly, with four checks.
A message sent to seed mailboxes that we read back ourselves, to see whether it reached the inbox or spam.
A placement testing tool sends from the mailbox to seeds at several providers.
Each list is first queried on a control address, to be sure it actually answers.
What mailbox providers say they receive in your name, authenticated or not.
How do you decide a mailbox can send?
Our rule is simple: a mailbox sends only on what we observed ourselves. A recent test in the inbox, no blacklist, authentication that passes. A score shown by a tool is not enough.
A check with no answer is not a pass
A test that could not conclude (a blacklist that does not answer, a seed message that never arrived) is recorded as unknown, never as "passed". Problems often hide in those gaps.
Modest volumes
We stay around 5 sends per day per mailbox. To send more, we add mailboxes and domains rather than pushing one address. A new address ramps up with real sends, slowly.
A check every week
Reputation moves. A domain that was healthy in September can slip in October. One test per week per domain is enough to see it before your prospects do.
Should you stop warmup altogether?
Not necessarily. A light warmup can help a brand new address break its silence, provided you watch your host and never read its score as proof of deliverability. The proof is placement tests.
Deliverability is only one part of prospecting: it lets the message be read, not be relevant. For what comes next, read our guides to B2B prospecting and sales prospecting.
Frequently asked questions
What is email warmup?
A service that has mailboxes in its own network send, open and rescue each other's messages from spam, to gradually raise the volume of a new address.
Do you need warmup for cold email?
No. A new address should ramp up slowly, but that can be done with modest real sends. What matters is checking where your emails land, with placement tests.
Why is my warmup score good while my emails go to spam?
The score measures exchanges inside the tool's network. Gmail judges your domain and IP reputation on its own traffic, which warmup does not reproduce.
How many cold emails per day per mailbox?
Few. We stay around 5 sends per day per mailbox, and add mailboxes rather than pushing the volume of one.
Every day Romeo spots the companies that need you and writes the first message. Nothing leaves without your approval.
Join the list