IP Warming in 2023: A 30-Day Schedule That Actually Works

Most IP warming schedules fail because they ignore engagement segmentation. The 30-day ramp that actually works volume tables, segment ordering, and recovery rules.

TL;DR

IP warming is the process of gradually ramping volume on a new sending IP so mailbox providers build positive reputation before the IP starts handling full traffic. Most published warming schedules fail because they treat warming as a volume problem when it is actually an engagement problem. The 30-day schedule that works: send only to your most-engaged segment in week 1 (top 10%), expand to top 25% in week 2, top 50% in week 3, and full volume by week 4. The volume table is secondary to segment ordering a 50,000-message campaign to your most engaged users on day 5 is safer than a 5,000-message campaign to your full list. Skip the warming and your IP will throttle for weeks. Skip the segmentation and you will warm the IP into a bad reputation that takes longer to recover than it would have to do correctly.

What IP Warming Actually Does

A new sending IP has no reputation. Mailbox providers do not know whether it belongs to a legitimate sender or a spammer who just rented a fresh IP block to bypass existing reputation. The default treatment is suspicious: throttle delivery, route some mail to spam, watch how recipients respond.

Warming is the process of demonstrating, through real sending behavior, that the IP belongs to a sender whose mail recipients want. You do this by sending only to recipients who will engage positively open, click, reply, not complain and ramping volume slowly enough that mailbox providers have time to update their reputation models.

The mechanism is straightforward. The execution is where senders fail.

The two most common failure modes:

1. Volume-only warming. Many published schedules show a daily volume ramp (5,000 day 1, 10,000 day 2, 20,000 day 3, etc.) without specifying who receives the mail. Senders interpret this as license to mail their full list at the prescribed daily volume. The result is a slow ramp delivered to a mix of engaged and unengaged recipients. Engagement rates look mediocre. Reputation builds slowly or not at all.

2. Skipping warming entirely. A sender migrates ESPs and assumes that "we have always sent 500,000 a day, why would we need to warm?" Volume is held constant, the new IP appears the next morning, and the first day's send hits a sub-50% inbox placement rate. Recovery takes 2-4 weeks of disciplined sending. The migration ends up costing more time than the warming would have.

Both failure modes have the same root cause: treating warming as something that happens to the IP independent of the audience. It does not. Warming is a co-evolution of IP reputation and recipient response.

Why Engagement Segmentation Matters More Than Volume

Mailbox providers measure reputation through observable behavior. The signals they care about most:

Open rate
Click rate
Reply and forward rate
Move from spam to inbox
Move from inbox to spam (negative)
Mark as spam (highly negative)
Mail sat unread for extended periods (slight negative)

When a new IP starts sending, these signals are the only data the provider has. A campaign sent to your most engaged 10,000 recipients produces dramatically better signals than a campaign sent to a random 10,000 from your full list. The difference, on a typical B2C marketing list, is roughly:

Most-engaged segment: 35-50% open rate, 5-8% click rate, ~0% complaint rate
Random segment: 15-25% open rate, 1-3% click rate, 0.05-0.15% complaint rate

To a mailbox provider scoring a new IP, those two profiles look like completely different senders. The first is a legitimate brand with a healthy audience. The second is ambiguous.

This is why segment ordering matters more than volume schedule. Volume too low: warming takes longer. Volume too high: warming damages the IP. Segment wrong: the volume does not matter the IP will not warm to a clean reputation regardless of pacing.

The 30-Day Schedule

The schedule below assumes a typical commercial sender migrating to a new IP, with reasonable list hygiene already in place (see List Hygiene Beyond Bounces for why this is a prerequisite). Adjust volumes proportionally for senders larger or smaller than the example.

Week 1: Top 10% Most Engaged Only

Audience: Recipients who have opened or clicked within the last 14 days. This is your warmest possible segment.

Volume:

Day 1-2: 1,000-2,000 messages
Day 3-4: 5,000 messages
Day 5-7: 10,000 messages

Goal: Establish a baseline of positive engagement signals on the new IP. Throttling at this stage is normal mailbox providers are deliberately limiting acceptance to observe behavior. Do not interpret slow delivery as a problem; it is the warming working.

What to send: Your highest-performing content. The campaigns that historically had the best engagement rates. Promotional content with a clear value proposition, not generic newsletters. The best-engaged segment receiving the best-performing content produces the strongest possible signals during the most critical phase.

Week 2: Top 25% Most Engaged

Audience: Recipients who have opened or clicked within the last 60 days. This expands the segment but stays well within the engaged audience.

Volume:

Day 8-9: 25,000 messages
Day 10-11: 50,000 messages
Day 12-14: 75,000-100,000 messages

Goal: Continue building positive reputation while increasing volume. By the end of week 2, the IP should be showing visible reputation in Postmaster Tools (if Gmail is a significant portion of your sending). For more on what to look for in Postmaster Tools.

Watch for: Any spike in complaint rate above 0.1% on the new IP. Any unexplained throttling that is not improving day over day. These are signs the segmentation is too aggressive or the IP has hit an underlying reputation problem (rare but possible if the IP was previously used by a spammer and hasnt been cooled of for long enough).

Week 3: Top 50% Most Engaged

Audience: Recipients who have opened or clicked within the last 90 days, or new subscribers from the last 30 days.

Volume:

Day 15-17: 150,000 messages
Day 18-21: 250,000 messages

Goal: Approach a substantial fraction of normal volume while staying on segments that engage. The newer subscribers in this segment matter they have not yet shown disengagement, and they tend to engage at higher rates than mature subscribers.

By the end of week 3, the IP's reputation should be stable and visible. Continued throttling at this stage suggests the underlying list quality is the problem, not the warming pace.

Week 4: Full Volume

Audience: Full active list (recipients within the 90-day engagement window).

Volume:

Day 22-25: Ramp toward full volume in 25% increments
Day 26-30: Full normal volume

Goal: Complete the warming and transition to steady-state sending. By day 30, the IP should be operating without warming-related restrictions. Inbox placement should match or exceed the previous IP's performance.

When 30 Days Is Not Enough

The 30-day schedule above is the typical case. Some scenarios require longer:

1. High-volume senders (>1M/day). The proportional ramp at the high end is large enough that volume jumps between days strain mailbox providers. Extend to 45-60 days, splitting the latter weeks into smaller daily increments.

2. Senders with poor historical engagement. If your full list has a 12% open rate, your "most engaged" segment is not very engaged compared to industry benchmarks. The warming signals will be weaker. Extend the schedule by 1-2 weeks and consider running a list hygiene pass before warming begins.

3. Senders with mixed transactional and marketing on the same IP. If the new IP will carry both streams, the warming has to work for both. Transactional warms faster (every message is opened) but the marketing volume is what stresses the IP. Extend by a week and stagger the ramp warm the transactional stream first, then layer marketing on top.

4. Migration from a damaged previous IP. If you are migrating because the previous IP had reputation problems, the receiving IP starts at zero but inherits some negative association from the sending domain. The domain reputation affects how mailbox providers interpret the new IP's behavior. Extend to 45 days and tighten the segmentation in the early weeks.

When 30 Days Is More Than Enough

A few scenarios where the schedule is overkill:

1. Transactional-only IPs. Receipts and password resets are opened at near-100% rates. A transactional IP can warm in 7-10 days because every message produces strong positive signals.

2. Senders with very high engagement ratios. Subscription-driven content (paid newsletters, premium products) often has 60%+ open rates and minimal complaint rates. These senders can warm in 14-21 days because the signals build faster.

3. Replacement IPs in the same pool. If you are adding capacity to an existing pool that already has good reputation on the related IPs in the same /24 block, the new IP often warms in 1-2 weeks because mailbox providers extend some trust based on the surrounding pool reputation.

What to Do When Warming Goes Wrong

The two most common warming problems and their resolutions.

Problem 1: Throttling does not improve after week 2.

Possible causes:

Segment is not as engaged as expected. Verify open and click rates on the warming sends are above 30% and 5% respectively.
The IP has prior negative reputation. Check the IP's history with reputation services like Talos Intelligence or Spamhaus. If the IP appears on blocklists, contact the ESP to either remediate or replace.
Underlying authentication is broken. Verify SPF, DKIM, and DMARC alignment on the new IP.

Resolution: Pause the schedule. Diagnose the root cause. Resume with the same week's volume only after the underlying issue is fixed. Do not advance to the next week's schedule while problems persist.

Problem 2: Complaint rate spikes during warming.

Possible causes:

Segment definition is too broad. The "engaged" criteria captured recipients who are actually disengaged.
Content quality is below the segment's expectations. Promotional content sent to recipients who only engage with high-value content produces complaints.
The previous IP's bad reputation has educated recipients to mark this sender as spam reflexively.

Resolution: Tighten the segment immediately. Drop to the previous week's audience definition. Run for 3-5 additional days at that tighter audience before resuming the schedule.

For senders consistently hitting complaint rate problems during warming, the fundamental issue is usually list quality, not warming execution.

A Concrete Example: Mid-Sized SaaS Migration

A SaaS company with 800,000 active marketing subscribers migrates from one ESP to another. The previous ESP had reputable shared IPs; the new setup uses a dedicated IP. Daily volume in steady state: 300,000-400,000 messages.

Pre-migration (week 0):

Audit list hygiene. Suppress everyone inactive 365+ days. List drops to 720,000 active.
Configure SPF, DKIM, DMARC on the new sending domain. Verify alignment.
Define engagement segments by query: opened/clicked in last 14 days (top 10%), in last 60 days (top 25%), in last 90 days (top 50%).

Week 1:

Day 1: 1,500 messages to top 10% (most-engaged). Subject line: best-performing recent campaign.
Day 3: 5,000 messages, same segment, different campaign.
Day 5: 10,000 messages, same segment.
Day 7: 10,000 messages, same segment.

Week 2:

Day 8: 25,000 messages to top 25%.
Day 10: 50,000 messages, same segment.
Day 12: 75,000 messages, same segment.
Day 14: 100,000 messages, same segment.

Week 3:

Day 15: 150,000 messages to top 50%.
Day 18: 200,000 messages, same segment.
Day 21: 250,000 messages, same segment.

Week 4:

Day 22: 280,000 messages to full active list.
Day 25: 320,000 messages, full list.
Day 28: 360,000 messages, full list.
Day 30: 400,000 messages, full list. Warming complete.

By day 30, the IP has positive reputation visible in Postmaster Tools, complaint rate is steady at 0.05-0.08%, inbox placement matches or exceeds the previous setup. The migration was operationally invisible to recipients.

Frequently Asked Questions

Do I need to warm a new IP if I am keeping the same sending domain?

Yes. IP and domain reputation are tracked separately. Even with established domain reputation, a new IP carries no reputation and will be throttled until it warms. The warming may go faster because of the domain reputation halo, but it cannot be skipped.

What if my ESP handles warming automatically?

"Automatic warming" usually means the ESP throttles your sends to match a volume schedule. This handles the volume side of warming but does not address segmentation. Verify that your sends during the automatic warming period are going only to engaged segments. If the ESP is throttling but you are still queueing your full list at the throttled rate, you are getting volume warming without engagement warming.

How long does warming take for a domain instead of an IP?

Domain warming follows similar principles but the timeline is shorter. New sending domains typically warm in 14-21 days because domain reputation builds faster than IP reputation in modern filtering models.

Can I warm two IPs simultaneously?

Yes, with caveats. Each IP needs its own engagement signal stream, so you cannot send the same campaign to the same recipient on both IPs and expect separate reputation. Split your audience into two non-overlapping halves and warm each on one IP. This is more complex than serial warming and not recommended unless you genuinely need parallel capacity.

What if I see no improvement after 30 days?

Either the segmentation was wrong, the underlying authentication is broken, or the IP has a prior reputation problem you did not know about. Review all three before extending the warming schedule.

Does Microsoft handle warming differently from Gmail?

Yes. Microsoft's filtering relies more heavily on domain reputation and SNDS-tracked IP reputation, both of which evolve more slowly than Gmail's engagement-based signals. Warming for Microsoft-heavy audiences benefits from a longer schedule (45+ days) and explicit attention to SNDS data during the ramp.

Can I run real campaigns during warming, or do I need to send "test" campaigns?

Real campaigns to engaged segments are the warming. There is no concept of warming sends being separate from production sends. Warming uses real sends to real recipients, just sequenced and segmented to produce strong signals.

Key Takeaways

IP warming is about engagement signals, not volume schedules. Segment ordering matters more than the daily volume table.
The 30-day schedule: top 10% most engaged in week 1, top 25% in week 2, top 50% in week 3, full volume by week 4.
Engagement criteria define the warming segments. Recipients who have opened or clicked within tight windows produce the strongest possible signals.
Throttling during warming is the system working, not failing. The IP is being observed, not penalized.
Warming cannot be skipped, even when migrating between ESPs with established reputation. New IP = no IP reputation, regardless of domain history.
When warming goes wrong, the cause is almost always list quality or authentication, not pacing. Diagnose before extending the schedule.
Transactional-only IPs warm faster (7-10 days). High-volume senders need longer (45-60 days). Most cases are 30 days.

The senders who skip warming pay for it later in delayed delivery and weeks of degraded inbox placement. The 30 days you spend warming correctly are an investment. Take them.

Reactions

ShareLinkedIn

Deliverability Strategy

Deliverability Incident Response: A Playbook for the First 4 Hours

When email reputation collapses, the first 4 hours determine the recovery timeline. Triage, containment, communication the operational playbook teams actually need.

Nov 26, 2024

Strategy Deliverability User Experience

List Hygiene Beyond Bounces: Engagement-Based Suppression Strategies

Bounce-based list cleaning is decades behind modern deliverability. The 90/180/365-day engagement model, sunset policies, and re-engagement that actually works.

Nov 21, 2024

Deliverability Regulation & Governance Security

Blocklist Remediation Playbook: Spamhaus, SORBS, and UCEPROTECT

How to remove your IP or domain from Spamhaus, SORBS, UCEPROTECT, and other blocklists. Which lists matter, which to ignore, and the delisting process for each.

Sep 17, 2024

Most IP warming schedules fail because they ignore engagement segmentation. The 30-day ramp that actually works volume tables, segment ordering, and recovery rules.

TL;DR

What IP Warming Actually Does

The mechanism is straightforward. The execution is where senders fail.

The two most common failure modes:

Why Engagement Segmentation Matters More Than Volume

Mailbox providers measure reputation through observable behavior. The signals they care about most:

Open rate
Click rate
Reply and forward rate
Move from spam to inbox
Move from inbox to spam (negative)
Mark as spam (highly negative)
Mail sat unread for extended periods (slight negative)

Most-engaged segment: 35-50% open rate, 5-8% click rate, ~0% complaint rate
Random segment: 15-25% open rate, 1-3% click rate, 0.05-0.15% complaint rate

To a mailbox provider scoring a new IP, those two profiles look like completely different senders. The first is a legitimate brand with a healthy audience. The second is ambiguous.