Reputation and IP warming series. This is the tactical guide to warming up a new sending IP. For related guides on domain reputation, warmup calculation, and cold IP recovery, see the reputation and warming hub →
Quick warmup reference
A new sending IP has no reputation. Receivers treat unknown IPs conservatively: rate-limit, greylist, or spam-fold until the sending pattern proves legitimate. IP warmup is the deliberate ramp from zero volume to full production volume over 4 to 8 weeks, starting with your most engaged recipients and expanding outward. Skip it and your first campaign hits deliverability walls that take months to recover from.
| Duration | 4 weeks for <100k target volume; 6-8 weeks for 100k+ |
| Growth pattern | Roughly 30-50 percent daily increase, capped by receiver tolerance |
| Segment order | Highest engagement first, then 30d engaged, then 60d, then 90d, then full list |
| Provider-specific limits | Gmail tolerates fast ramp; Microsoft is strictest; Yahoo mid-range |
| Trigger to slow down | Bounce rate > 2 percent, complaint rate > 0.1 percent, engagement dropping |
| 2026 context | Gmail and Yahoo weight domain reputation more than IP now, but IP warmup still matters for dedicated IPs and shared-IP pools |
Who this guide is for
- Sender launching a new dedicated IP from cold
- Deliverability team migrating from shared to dedicated
- Ops team preparing for 2024 Gmail and Yahoo bulk sender compliance
What you’ll learn
- The exact 30/60/90 day daily volume ramp
- Segment selection by engagement tier (7d, 30d, 60d, 90d)
- KPI checkpoints per week and provider-specific caps
The reputation and warming series (4 guides)
1. IP warmup schedule
READINGThe day-by-day ramp for new IPs
2. Domain vs IP reputation
Which score matters where in 2026
3. Warmup calculator
Formulas, Python and tables per target
4. Cold IP recovery
When reputation crashed and you rebuild
IP warmup is one of those disciplines where the difference between doing it right and doing it approximately is measured in months of deliverability. Warm too fast and receivers throttle you into oblivion; warm too slow and you waste weeks before hitting production volume. The right approach follows a specific curve, with specific segment selection at each step, and specific KPI checkpoints that tell you to slow down before damage becomes irreversible. This guide covers the exact 30/60/90 day ramp, how to pick which recipients to send to at each stage, what to watch, and what to do when warmup goes sideways.
Why cold IPs need warmup
A brand new sending IP has no history at any receiver. Gmail, Microsoft, Yahoo, and every other major receiver keep per-IP reputation records: how much this IP sent, how much was accepted, what the bounce and complaint rates looked like, what fraction of recipients engaged. A new IP has none of that. The receiver’s default behavior for unknown-IP mail:
- Rate-limit acceptance. Cap the number of messages per hour or per day from the new IP until reputation builds.
- Greylist. Reject the first attempt with a temporary 4.7.1 code, forcing a retry that proves the sender is a legitimate MTA (not a spam botnet).
- Bias toward spam folder. Even accepted messages get delivered to spam or promotions until engagement signals prove otherwise.
If your first campaign is 500,000 messages from a brand new IP, most of them fail delivery, and the receivers now have data saying “this IP sends huge unwanted campaigns.” That reputation damage persists for weeks. The warmup process avoids this by proving to receivers, over time, that your sends are wanted (engagement) and technically correct (low bounce and complaint rates) before you hit production volume.
When you need to warm up (and when you don’t)
You need warmup if
- You are moving to a new dedicated IP.
- You are launching a new sending domain, even on a shared IP.
- You are recovering from a reputation crash (see the cold IP recovery guide).
- You have paused sending for 30+ days and are re-entering.
You do not need full warmup if
- You are joining a shared IP pool that already has established reputation (typical ESP starter tier). The pool’s reputation carries you.
- You are adding a new domain to an existing warm IP with a strong sending history.
- Your total monthly send volume is under 10,000 messages (below the threshold where receivers care much about reputation).
The 30/60/90 day ramp: exact numbers
The classic ramp for a target volume of ~100,000 messages per day, adapted to 2026 receiver behavior:
Week 1 (Days 1-7): Foundation
| Day | Volume | Segment |
|---|---|---|
| 1 | 50 | Top 5 percent engaged (opened last 7 days) |
| 2 | 100 | Same segment |
| 3 | 500 | Same segment |
| 4 | 1,000 | Expand to top 10 percent engaged |
| 5 | 2,500 | Same segment |
| 6 | 5,000 | Same segment |
| 7 | 7,500 | Expand to top 20 percent engaged |
Week 2 (Days 8-14): Momentum
| Day | Volume | Segment |
|---|---|---|
| 8 | 10,000 | Top 20 percent engaged |
| 10 | 15,000 | Expand to top 40 percent engaged |
| 12 | 25,000 | Same |
| 14 | 35,000 | Expand to top 60 percent |
Week 3 (Days 15-21): Widening
| Day | Volume | Segment |
|---|---|---|
| 15 | 45,000 | Top 60 percent |
| 17 | 55,000 | Expand to 30-day engaged (top 80 percent) |
| 19 | 65,000 | Same |
| 21 | 75,000 | Same |
Week 4 (Days 22-30): Full volume
| Day | Volume | Segment |
|---|---|---|
| 22-25 | 85,000 | 90-day engaged |
| 26-30 | 100,000 | Full active list |
For larger target volumes (500k+ per day), extend the ramp to 6-8 weeks by doubling the number of days at each tier. For smaller targets (<10k per day), compress to 2 weeks but keep the same segment expansion order.
Real-world warmup scenarios: which ramp fits you
SaaS B2C, 100k target, engaged consumer list
STANDARDFollow the 30-day ramp exactly. Growth rate 40 percent daily. Start with 7-day engaged (top 5-10 percent), expand to 90-day engaged by end of week 3.
Duration: 28-30 days · First send: 50 recipients · Watch: Postmaster Domain reputation daily
B2B startup, 15k target, opt-in professional list
COMPRESSEDCompress the ramp to 14-21 days. Growth rate 35 percent daily. B2B engagement is naturally lower, so oversize the engaged segment slice and start smaller (25 recipients).
Duration: 14-21 days · First send: 25 recipients · Watch: Microsoft SNDS (many B2B on Outlook 365)
Reactivation after 90+ day pause on the same IP
RE-WARMThe IP has decayed reputation (not zero, but not warm). Growth rate 25-30 percent daily. Start with last 30-day engaged only. Reputation systems remember; go slower than a fresh warmup.
Duration: 21-30 days · First send: 200 recipients · Watch: Sender Score trend (rebuild vs decay signal)
Segment selection: engagement is the signal
The volume numbers are only half the plan. The other half is which recipients you send to at each stage. Warmup works because you start with recipients most likely to engage positively (open, click, not complain), and receivers see those positive signals and build reputation upward.
Engagement tiers (rank recipients by these, top to bottom)
- 7-day openers. Opened at least one message in the last 7 days.
- 30-day openers. Opened at least one message in the last 30 days but not in the last 7.
- 60-day openers. Between 30 and 60 days ago.
- 90-day openers. Between 60 and 90 days ago.
- Older openers. Beyond 90 days but ever opened.
- Never-engaged. Subscribed but never opened. Consider excluding entirely; these are usually spamtraps or dead addresses.
Selection rules per stage
- Week 1: only 7-day openers. Highest engagement, lowest complaint risk.
- Week 2: add 30-day openers.
- Week 3: add 60- and 90-day openers.
- Week 4: full active list (last 12 months of engagement).
- Never-engaged: exclude from warmup entirely. If you must include them, do so after full warmup, in small batches with tight monitoring.
Provider-specific tolerance
Gmail
Most tolerant of fast ramps if the engagement is real. Gmail’s algorithms weight opens and click behavior heavily. A ramp that shows steadily rising engagement over the first two weeks can accelerate somewhat. Watch the Spam Rate in Google Postmaster Tools daily; if it stays under 0.1 percent, you have headroom.
Microsoft (Outlook.com, Hotmail, Live)
Strictest of the majors. Microsoft’s SmartScreen algorithms are conservative about new IPs and slow to lift throttling. Expect Outlook delivery to lag Gmail by 1-2 weeks. Sign up for JMRP (Microsoft’s FBL, see our FBL guide) and SNDS. Do not exceed 25 percent of daily volume to Outlook addresses in weeks 1-2.
Yahoo (and AOL, Verizon)
Mid-range tolerance. Yahoo requires DKIM alignment and gets touchy about volume spikes. Register for Yahoo’s Complaint Feedback Loop (see FBL guide). Keep Yahoo volume increases modest across the ramp; sudden 2x days catch attention.
Apple iCloud
Opaque. No public FBL, no dashboard equivalent to Postmaster Tools. Watch aggregate delivery metrics (open rates) as an indirect signal. If iCloud open rates drop 30 percent below expected during warmup, slow down.
Smaller providers
Comcast, La Poste, Terra, UOL, and regional ISPs are mostly permissive of new IPs but strict on complaint rates. Follow the same ramp; register for the FBLs they offer.
KPI checkpoints per week
Before advancing to the next week’s volume, verify these metrics on the previous week’s sends. If any metric fails, hold at current volume for another 3 days and re-evaluate.
| Metric | Target | Action if failed |
|---|---|---|
| Bounce rate | < 2% | Hold volume; verify list hygiene |
| Hard bounce rate | < 0.5% | Suspend; verify recipient list quality |
| Complaint rate | < 0.1% | Hold volume; review content and targeting |
| Delivery rate | > 97% | Investigate blocklists, auth setup |
| Open rate | > 20% (varies by industry) | Segment quality issue; hold and re-verify engagement tier |
| Postmaster Tools reputation | Rising or High | Hold if Medium; suspend if Low |
A single bad day is not a signal to slow down; a 3-day trend is. Do not overreact to normal variance.
What breaks warmup (and how to recover mid-ramp)
Sudden volume spikes
Doubling volume in a single day, even if within the planned ramp, spikes receiver alerting. Smooth increases across days.
Content changes mid-warmup
Changing the sending pattern (from newsletters to promotional to transactional) during warmup confuses reputation systems. Send the same type of content throughout the ramp; introduce diversity only after warmup completes.
Bad list quality
Even a high-engagement segment can contain dead addresses or spamtraps if the list was purchased or scraped. Verify the warmup segment with an address validation tool before starting; a 2 percent bounce rate in week 1 wrecks the entire ramp.
Blocklist listing
New IPs occasionally get preemptively listed by SORBS or Spamhaus if the /24 block has bad neighbors. Check on day 1 and periodically at blocklist directory; if listed, follow delisting procedures immediately.
Recovery: hold and re-verify
If any KPI fails, do not roll back. Hold at the current volume for 3-5 days, verify the underlying issue, then resume the ramp when metrics stabilize. Rolling back mid-warmup sends conflicting signals to receivers.
10 common IP warmup mistakes
- Warming with the full list from day 1. Sending to disengaged recipients too early spikes bounce and complaint rates.
- Skipping warmup for “urgent” campaigns. A blocked IP is more urgent than a delayed campaign. Warmup is not optional.
- Warming to only Gmail. Microsoft and Yahoo need separate signals. A warm IP at Gmail is still cold at Outlook.
- Ignoring Postmaster Tools during warmup. The dashboard shows problems days before they become deliverability crashes.
- Continuing warmup during a metric failure. Bounce rate spikes to 3 percent, but the schedule says increase, so you increase. This is how warmups turn into disasters.
- Warming with never-opened recipients. Even “old subscribers” who never opened are the worst candidates for warmup. They are spamtrap risks.
- No DKIM alignment before warmup. DKIM-unaligned mail hurts reputation from day 1. Configure SPF, DKIM, DMARC before starting; see our authentication pillars.
- Sending multiple content types during warmup. Mixing newsletter + promotional + transactional in the same warmup period sends conflicting reputation signals.
- Not correlating with per-provider volume. If 60 percent of your list is Gmail, do not send 60 percent of your day-1 volume to Gmail; split proportionally and never exceed the per-provider cap.
- Ending warmup prematurely. Reaching target volume on day 28 does not mean warmup is done. Sustained low complaint and high engagement for another 2 weeks is what confirms the ramp worked.
IP warmup FAQ
How long does IP warmup actually take?
Between 30 and 90 days depending on target volume and receiver mix. Small senders (target under 50,000/day) can finish in 30 days if metrics stay green. Large senders (100,000+/day) typically need 45-60 days. Very high-volume senders (1M+/day) may take 60-90 days. Rushing the ramp to hit deadline is the most common cause of deliverability crashes; the metrics decide the pace, not the calendar.
Should I warm up multiple IPs in parallel?
Yes if you plan to run them as a pool. Each IP needs its own warmup because reputation is per-IP. Split your warmup segment across the IPs (each IP gets a slice) or warm them sequentially over overlapping windows. Parallel warmup takes the same calendar time as single-IP warmup but requires more total volume and more monitoring bandwidth. Sequential warmup is safer but takes longer to finish the whole pool.
Does IP warmup still matter in 2026 given the domain reputation shift?
Yes, though the weighting has shifted. Gmail and Yahoo now weight domain reputation more than IP reputation, and Microsoft continues to weight IP significantly. If you use a dedicated IP or a small shared-IP pool, warmup remains important. If you use a large shared pool at a major ESP (SendGrid, Mailgun, Postmark shared), the pool has established IP reputation and warmup is minimal; domain reputation becomes the primary focus. See our next article on domain vs IP reputation.
Can I use warmup services like Warmup Inbox or Mailwarm?
These services simulate engaged sends by having a network of accounts open and click your test messages. They can help build IP and domain reputation faster in early warmup, particularly for cold email senders who do not have a real engaged list. Two caveats: (1) receivers are increasingly good at detecting artificial engagement and may discount it; (2) they do not substitute for eventually sending real content to real recipients. Useful as a supplement in the first 1-2 weeks, not a replacement for the full ramp.
What if I hit a metric failure at week 3?
Hold at current volume for 5-7 days. Investigate the specific KPI that failed: bounce rate suggests list quality, complaint rate suggests content or targeting, delivery rate suggests infrastructure (auth, blocklists). Fix the underlying problem before resuming the ramp. Do not roll back to lower volumes; that sends conflicting signals. If the problem is severe (bounce rate 5%+ or complaint rate 0.3%+), pause entirely for a week and consult your reputation dashboard before resuming.
Should I include suppression list checks during warmup?
Always, at every stage. The suppression list check is non-negotiable regardless of warmup state. A recipient who complained on a previous IP is even more likely to complain on your new IP. Pre-flight check every send in warmup exactly as you would at full production volume.
Final words
IP warmup is a discipline that rewards patience and punishes shortcuts. The 30/60/90 day framework works because it lets receivers observe increasing volume alongside stable engagement metrics, which is exactly the pattern of a legitimate sender growing organically. Skip steps and receivers see the pattern of a spam campaign scaling up too fast; the result is the same rate-limits, spam-folding, and reputation damage that took weeks to earn.
The mental model: warmup is a conversation with receivers, not a broadcast. Every day, you send a certain volume; every day, the receivers watch how recipients react; every day, the reputation math updates. Give receivers time to build a positive reputation model before you ask them to accept full-scale sends.
For the follow-up articles in this series: domain vs IP reputation covers the 2026 shift in receiver weighting, the warmup calculator gives you the exact formula to compute your ramp, and cold IP recovery covers the case where reputation crashed and you need to rebuild.
For related content, see the best email warmup tools review, the foundational warm-up guide, Sender Score, Google Postmaster Tools, the Google and Yahoo bulk sender rules, and the Deliverability hub.
A clean list is the foundation of every warmup.
SMTPing catches invalid addresses, disposables, catch-all traps, role accounts and dead mailboxes before your first warmup send. A 2 percent bounce rate on day 1 kills your ramp. 13 validation types, 25 free daily.
About the Author
Alaa · LinkedIn
Email infrastructure specialist with 8+ years of hands-on experience in SMTP, deliverability, and email verification. I’ve configured and troubleshot mail systems across Postfix, Exchange, and cloud relays, managed IP reputation and warmup campaigns, and built verification pipelines processing millions of addresses. My work spans DNS authentication (SPF, DKIM, DMARC, BIMI), bounce handling, blocklist monitoring, and compliance frameworks including CAN-SPAM and GDPR. I write every article on SMTPedia to give email professionals, developers, and marketers the accurate, RFC-grounded reference they need.
About SMTPedia
SMTPedia is an independent email industry reference covering SMTP, IMAP, POP3, email deliverability, marketing platforms, DNS authentication, and email verification. Every article is researched from official provider documentation, IETF RFCs, and industry best practices. Settings and configurations are verified quarterly.
We are cited as a source by ChatGPT, Microsoft Copilot, and thousands of email professionals worldwide. Learn more about our editorial process.
