There is a moment in every email programme where you commit a small, irreversible mistake: you start sending at full volume from an IP or domain nobody has heard of.
Mailbox providers do not know you. They have no record of your sending pattern, no history of recipients opening your mail, no complaints on file for you either way. And then forty thousand messages arrive at once from a source with no track record.
To a filtering system, that is indistinguishable from a spam run. Your first campaign goes to spam, and the recovery is measured in weeks.
Warmup is the deliberate avoidance of that outcome. It is the process of earning a sending reputation the slow, boring way — and it is the single most skipped deliverability discipline in email marketing.
IP warming is the practice of gradually increasing send volume from a new or recovered IP or sending domain, so mailbox providers accumulate evidence that your mail is wanted. Because an unfamiliar sender has no reputation, sending large volumes immediately reads as spam behaviour.
This guide covers what providers evaluate, a phased warmup schedule, which subscribers to contact first, the metrics that tell you to hold or increase, and the mistakes that burn a domain in the first week.
What you will learn:
- What mailbox providers actually evaluate when you are new
- A phased warmup schedule you can follow immediately
- Which subscribers to send to first, and why order matters
- The metrics that tell you to hold or increase volume
- The mistakes that burn a domain in the first seven days
The Short Version
- A new domain or IP has no reputation, which providers read as risk.
- Sending at volume immediately reads as a spam run, so placement collapses.
- Warmup is a gradual ramp, typically three to six weeks to full volume.
- Start with your most engaged subscribers — engagement is what builds reputation.
- Increase volume only while placement and complaints stay healthy.
Direct answer: you cannot "switch on" warmup for a domain that has already sent cold. Warmup is a forward-looking ramp you start before your first campaign. If you have already sent at volume from a new domain, the recovery path is different and is covered under sender reputation recovery.
Why Warming Is Necessary
What Providers Actually Evaluate
Warming works because it addresses a specific problem in provider filtering: unfamiliar senders are unclassifiable, and unclassifiable senders are risky.
Mailbox providers weigh several signals continuously:
- Engagement with your specific recipients. Do people who receive your mail actually open and click it?
- Complaint rate. What proportion of recipients mark you as spam?
- Bounce rate. Are you sending to addresses that do not exist?
- Send pattern stability. Is volume predictable, or erratic?
- Authentication alignment. Do SPF, DKIM and DMARC pass consistently?
A new sender has no engagement history, no established send pattern, and — critically — an engagement record that starts at zero. If you send to a large, unengaged list immediately, you will generate opens and complaints on day one, and that early record becomes the foundation of your reputation.
Why Volume Spikes Trigger Punishment
Providers are explicitly tuned to detect bulk-send behaviour, which has two recognisable signatures: a large first send from an unknown source, and a sudden step change in volume. Both signal spam runs, which is why 47 reasons your email was rejected includes volume patterns among its causes.
This is also why warmup is not a formality. Providers now apply stricter thresholds to bulk senders than ever before, and the requirements are explicit in the Gmail and Yahoo bulk sender rules. Warmup is how you satisfy those thresholds from a standing start.
| Factor |
Cold sender |
Warmed sender |
| Reputation history |
None |
Established |
| Engagement signal |
Zero or unreliable |
Consistent and positive |
| Complaint rate |
Unknown, risky |
Low and monitored |
| Send pattern |
Unknown |
Predictable |
| Provider classification |
Unclassified / risky |
Known sender |
The Warmup Schedule
Here is a workable four-phase ramp. Treat it as a template rather than a fixed rule — the right volume for you depends on list size and engagement.
| Phase |
Volume |
Duration |
Who receives |
Advance when |
| 1. Pilot |
Low tens to low hundreds/day |
3–5 days |
Most engaged only — recent clickers, recent buyers |
Messages land in inbox, replies arrive |
| 2. Ramp |
Increase ~25–50% per day |
Week 1–2 |
Engaged plus active segments |
Placement stable, no complaint spike |
| 3. Expand |
Double roughly every 2–3 days |
Week 2–4 |
Full list, including less engaged |
Placement holding across providers |
| 4. Steady state |
Normal cadence |
Ongoing |
Full list |
Ongoing monitoring |
How to Think About Volume
Percentage of list size is a better mental model than an absolute number. A list of 5,000 does not need 500 emails a day to look legitimate; a list of 500,000 sending 500 a day is not warm in any meaningful sense.
Many experienced senders combine a percentage approach with a hard per-mailbox-per-domain cap, so volume can only rise as the ratio supports it.
Why Double Rather Than Jump
Each increase tells the provider "this sender's volume is predictable and rising gradually". A five-fold jump breaks that pattern and restarts the clock. Slow is genuinely faster here.
The email IP warmup planner will turn your specific list size into a dated schedule you can work from, which removes the most common failure mode: guessing at volume and ramping too fast.
Send Order Matters More Than Most People Expect
The most common warmup failure is not ramping too fast — it is ramping to the wrong audience.
Warmup is meant to accumulate positive engagement signals. If your first ten thousand sends go to the least engaged segment of your list, you will generate almost no engagement, some complaints, and a reputation built on the weakest possible foundation. You are not warming; you are establishing that nobody wants your mail.
Send in reverse order of engagement quality:
- Recent purchasers — strongest signal available, and useful post-purchase content.
- Recent clickers — demonstrated interest, easy to re-engage.
- Active site visitors — engaged, even if they never opened your mail.
- Recent opt-ins — new, motivated, and expecting contact.
- Older openers, no recent clicks — workable, but thinner signal.
- Dormant subscribers — last, if at all. These belong in a re-engagement flow, not a warmup.
Segment 6 deserves a specific warning. Sending to dormant subscribers during warmup is how well-meaning senders manufacture a complaint problem while trying to build reputation. If you have a large inactive segment, exclude it and work it through a separate re-engagement campaign — where a winback calculator can tell you whether it is worth doing at all.
A Worked Example: Warming a 50,000-Subscriber List
A retailer has 50,000 subscribers, mostly from purchases over the last year. They have a new dedicated IP and no reputation. Here is how the ramp would run.
Phase 1 — Pilot (days 1–5). 100 emails a day, all to recent purchasers. That is 0.2% of the list. Expect roughly 90% inbox placement and some replies. If placement is below 90% at any point, stop and fix authentication rather than continuing.
Phase 2 — Ramp (days 6–12). Increase about 30% per day: 130, then 170, then 220, then 280, and onward past 500. Segments added: recent clickers, then active site visitors. You are now at roughly 700 a day, about 1.4% of list.
Phase 3 — Expand (days 13–25). Double every two to three days, reaching roughly 8,000 a day by day 25. Segments added: recent opt-ins, then older openers. The dormant segment — perhaps 15,000 subscribers — stays excluded.
Phase 4 — Steady state (week 4 onward). The dormant segment is worked through a separate winback campaign, and normal cadence begins at full volume.
What makes this work: volume stays low for most of the ramp. It is under 1.5% of list size for the first twelve days and only reaches roughly 16% at the very end, increases are gradual throughout, and the least engaged audience is contacted last rather than first.
The Metrics That Tell You to Hold or Increase
Warmup without measurement is just slow sending. Four signals tell you whether to advance.
| Metric |
Healthy during warmup |
Warning sign |
| Inbox placement rate |
Stable or improving |
Declining as volume rises |
| Spam complaint rate |
Low and flat |
Any upward trend |
| Bounce rate |
Under ~2% |
Rising above 2–3% |
| Engagement |
Above your own baseline |
Below baseline despite more sends |
The inbox placement rate calculator is the most important of these, because placement is the outcome you are warming toward and the earliest indicator that something has gone wrong. A volume increase that drops placement was a mistake — reduce volume back to the last level that worked and hold there.
Watch the complaint rate just as closely. Warmup is specifically a period where providers are forming an impression, and complaints during this window are disproportionately costly. The spam complaint rate calculator will tell you whether you are inside or outside a safe threshold, and the deliverability loss calculator will quantify what a given placement rate is costing you in revenue.
One measurement trap to avoid here: do not judge warmup progress on open rate. Apple Mail records opens on delivery rather than reading, so an open rate can rise while you are being pushed toward spam — the exact opposite of what you want to see. Use placement, complaints and clicks, as covered in Apple Mail Privacy Protection.
Domain Warmup vs. IP Warmup
These are related but distinct problems, and conflating them causes real confusion.
Domain warmup is about establishing a sending identity that providers recognise and trust. A new domain has no history, so you build reputation gradually from a low base. This is the more important of the two, because a bad domain reputation can affect you across every IP you use.
IP warmup is about establishing reputation for a specific sending source. This matters most when you use a dedicated IP, which gives you a consistent identity and clean reporting but requires you to earn reputation from scratch.
| Consideration |
Domain warmup |
IP warmup |
| Affects reputation across all IPs? |
Yes |
No |
| Needed for a shared pool? |
Yes |
Controlled by the pool |
| Needed for a dedicated IP? |
Yes |
Yes |
| Longer to repair if damaged? |
Usually |
Sometimes faster |
A shared or reseller pool inherits the provider's existing reputation, which means you start warmer but with far less control and no isolation from other senders' behaviour. Dedicated IPs give you control and clean data but make you entirely responsible for your own ramp. The trade-offs are covered in IP reputation explained, and if you are currently blocked, sender reputation recovery covers the repair path.
Whichever you are warming, authentication is the baseline you cannot skip. SPF, DKIM and DMARC failures will block you no matter how carefully you ramp, so verify alignment before your pilot send rather than after. Start with email authentication methods explained and the authentication health check.
The Mistakes That Burn a Domain in Week One
Almost every failed warmup comes down to one of these.
- Launching to the full list on day one. The single most common cause of a failed warmup, and the least forgivable.
- Ramping by a large multiple. Tripling or quintupling volume in one step reads as a sudden pattern change and undoes prior progress.
- Sending to the dormant list first. Generates complaints during the exact window you are being judged on.
- Skipping authentication. DMARC failures are an immediate, unambiguous trust signal — in the wrong direction.
- Buying or scraping more addresses to "get more data". Volume is not the goal; engaged volume is. This is the same trap described in spam traps and list hygiene.
- Running a promotional blast during warmup. Deep discounts produce complaint-prone, low-quality engagement. Save the sale.
- Warmup and cold outreach on the same infrastructure. Contaminates the ramp. Keep them separated, as described in cold email vs. marketing email.
- Ignoring placement during the ramp. If you only check placement at the end, you will find out too late that day four went badly.
A Four-Week Execution Plan
Here is the sequence, in order, with nothing skipped.
Week 0 — Before Any Send
Confirm SPF, DKIM and DMARC are aligned and passing. Set up inbox placement seed testing. Define your segment order by engagement. Confirm you can measure complaints and placement per campaign. Calculate your target volume at each phase.
Week 1 — Pilot and First Ramp
Send only to purchasers and recent clickers at low volume. Confirm inbox placement by hand each day. Advance only while placement holds.
Week 2 — Expand to Engaged Segments
Increase roughly 25–50% per day. Begin incorporating opt-ins and active visitors. Hold volume if placement dips.
Week 3 — Full List
Send to everyone except the dormant segment. Watch the bounce rate closely, since this is where list quality problems surface. Adjust volume up or down daily based on placement.
Week 4 — Steady State and Monitor
Reach normal cadence. Do not stop measuring. Establish a baseline you can compare future campaigns against, using the email deliverability calculator and the email benchmark comparator.
Before any seasonal peak, run the same discipline in reverse — a controlled slowdown rather than an abrupt pause. Abruptly stopping and restarting is itself an erratic pattern. The Black Friday deliverability checklist covers this in more detail for high-volume periods.
Key Takeaways
- An unfamiliar sender is unclassifiable, which providers treat as risk.
- Warmup is a gradual ramp, typically three to six weeks.
- Start with your most engaged subscribers — engagement builds reputation.
- Increase volume only while placement and complaints stay healthy.
- Authentication failures block you regardless of how carefully you ramp.
- Seven common mistakes — including launching to the full list on day one — cause most failures.
Sources and Further Reading
Related Articles
Related tools: Generate a dated ramp with the email IP warmup planner, track the outcome with the inbox placement rate calculator, and confirm your standing with the sender reputation checker.
This article provides general guidance on email deliverability and is not a guarantee of inbox placement. Provider thresholds and filtering behaviour change over time, so measure your own program and verify current guidance.
