Every team that starts a serious cold outreach program asks roughly the same question in roughly the same week: "how long until we can send at full volume?" The honest answer is longer than they expect, and the reasons are structural rather than arbitrary. Inbox providers built their filtering around the assumption that legitimate senders establish reputation gradually, through engagement patterns that look like real human correspondence. Cold outreach violates that assumption by definition — your recipients did not ask to hear from you — which means your only defense against being filtered is to look, for as long as it takes, like exactly the kind of sender whose mail providers want to deliver. This article is about setting realistic expectations for that timeline, what you are actually buying with the patience, and why the teams that try to shortcut it end up starting over twice as often as the teams that accept the timeline on the first attempt.
Key takeaways
- Warmup is the process of establishing sender reputation on a new domain, a new mailbox, or a new IP before significant cold volume is sent. Without it, providers treat your mail as suspicious and filter accordingly.
- The realistic timeline is four to eight weeks from a standing start to full sending capacity. Teams that compress this to a week get away with it on the first campaign and pay the cost on the next three.
- Daily volumes during warmup ramp from single digits in week one to thirty to fifty per mailbox by weeks six to eight. These are per-mailbox caps, not total program caps — scale is achieved through mailbox count, not by pushing individual mailboxes past their sustainable ceiling.
- The engagement signals that matter are opens, replies, and positive folder actions (starring, moving out of spam, adding to contacts). Sends without engagement build no reputation regardless of volume.
- GDPR, in force since May of this year, has changed the calculus for cold email in Europe specifically. Legitimate-interest legal basis remains defensible for many B2B scenarios, but the operational hygiene around opt-out and data minimization has to be stronger than it was pre-GDPR.
Why warmup expectations are usually wrong
The gap between expectations and reality on warmup comes from two sources. The first is that the internal stakeholders asking the question — sales leadership, marketing leadership, the CEO wanting to know when pipeline will materialize — are not in the cold email operational loop. They see the product rolling out, they see the target list ready, and they assume the rate-limiting step is something under the team's control. They are surprised to learn that the rate-limiting step is provider-side, and that no amount of internal urgency speeds it up.
The second source is the marketing of cold email tools. Vendor messaging tends to emphasize features that shorten the apparent setup time — "launch your first campaign in 15 minutes," "automated warmup built in." Both claims are technically true and operationally misleading. You can launch a campaign in fifteen minutes. The campaign will perform poorly because the underlying sender reputation has not been built. Automated warmup helps, but it cannot compress the provider-side trust-building below its natural timeline; it can only make the process more consistent and less manual.
The realistic framing for leadership: cold email pipeline from a new domain starts showing up in four to six weeks after setup, with meaningful volume in six to ten weeks, and steady-state performance at the twelve-week mark. Any shorter timeline is either an existing-domain warmup (different starting point) or a forecast that will miss its numbers. Teams that set expectations with leadership accordingly avoid the "why isn't cold email working yet?" conversation in week two that often derails the program entirely.
What warmup is actually doing
Understanding why warmup takes the time it does requires understanding what the provider is observing on the other side. A new domain sending cold email presents a specific pattern: sending to recipients with no prior relationship, at volumes that ramp over time, with content that is broadly similar across recipients. The provider's job is to distinguish this pattern from the similar-looking pattern produced by spammers, and the distinction is built from engagement signals over time.
The signals the provider is watching
| Signal | What a legitimate warmed-up sender looks like | What a cold spammer looks like |
|---|---|---|
| Volume ramp | Starts low; ramps consistently over weeks | Large volume from day one, then drops |
| Recipient engagement | Opens, replies, positive folder actions | Low engagement; high spam complaints |
| Bounce rate | Low (under 2%) with clean list hygiene | High; lists are frequently dirty |
| Complaint rate | Very low (under 0.1%) | Elevated; recipients often mark as spam |
| Authentication | SPF, DKIM, DMARC all passing consistently | Partial or broken authentication common |
| Reply behavior | Real replies flow back through the same domain | Broken reply paths; no real conversations |
| Sending cadence | Consistent pattern; business hours; predictable windows | Bursty; unusual hours; erratic timing |
Warmup is the process of producing all of these signals consistently, across enough time, that the provider's classifier moves you from "unknown new sender" to "established legitimate sender." Nothing about that process is artificial; you are not fooling the classifier. You are demonstrating, through actual sending behavior, that you belong in the legitimate category. The classifier updates its estimate accordingly. The underlying model of how that classification works in the first place — what the providers are actually measuring — is covered in sender reputation fundamentals.
The realistic timeline from day one
The timeline below assumes a fresh outreach domain, fresh mailboxes, and a new team running the warmup. Teams with existing reputation on part of the stack can compress some of the phases, but the compressed path is the exception rather than the default.
| Phase | Duration | What happens | What you can send to cold prospects |
|---|---|---|---|
| 1. Setup | Week 0 (before warmup starts) | Domains registered; DNS published; SPF/DKIM/DMARC verified; mailboxes created and configured | Nothing. If you send cold mail during setup, you fail before you start. |
| 2. Initial warmup | Weeks 1-2 | 5-15 daily emails per mailbox, all to friendly recipients or warmup network; high engagement rates — the underlying ramp logic is the same one described in designing an IP warmup schedule, applied at the mailbox level | Nothing. Domain has no reputation yet. |
| 3. Gradual warmup | Weeks 3-4 | 15-30 daily per mailbox; engagement still dominant; start introducing mixed content | Very limited. 5-10 emails per day to high-quality prospects, only after week 3 |
| 4. Ramped outreach | Weeks 5-6 | 30-40 daily per mailbox; cold outreach can form majority of sends | Yes — 20-30 per day per mailbox to prospects |
| 5. Full volume | Week 7+ | 40-50 daily per mailbox sustained; warmup traffic continues at reduced percentage | Yes — at the sustainable per-mailbox cap |
| 6. Steady-state maintenance | Ongoing | 15-20% of daily volume remains warmup traffic indefinitely | Full production volume continues |
The absolute timeline from domain registration to full cold sending capacity is thus six to eight weeks minimum, with seven being the typical middle. Teams that compress this to three or four weeks get lucky sometimes; teams that compress it to two weeks rarely escape the consequences. The consequences compound: a domain rushed through warmup develops weaker long-term reputation than a domain warmed up properly, and the difference persists for the life of the domain.
Daily volume targets week by week
The specific per-mailbox daily volumes below are the practical ceilings for 2018 cold email programs. These numbers drift downward over time as providers tighten their filters; what works this year will be slightly lower next year. Plan conservatively.
Notice two properties of the curve. First, the total per-mailbox volume never exceeds the fifty-per-day ceiling; the growth is in the cold proportion, not in raw daily throughput. Second, warmup volume never hits zero — the steady-state pattern keeps ten to fifteen warmup emails per mailbox per day indefinitely, because that maintenance traffic is what keeps the mailbox's reputation healthy through content variations and quiet periods.
Engagement signals that make warmup work
Volume without engagement is not warmup; it is just sending that teaches the provider nothing positive. Every warmup email needs to generate engagement, and the specific types of engagement matter. The ranking below reflects what we observe in Google Postmaster Tools and SNDS as the strongest trust-builders.
- Reply. A recipient responding to the message is the single strongest positive signal. Any warmup that produces real replies is building reputation faster than one that produces only opens.
- Moving out of spam folder. If the recipient finds the message in spam and marks it as "not spam" or moves it to inbox, that is an explicit trust signal and extremely valuable.
- Starring or marking as important. Some recipients star or flag messages they consider significant. Providers weight this highly.
- Adding the sender to contacts. An explicit "add to contacts" action from the recipient's side tells the provider this sender is someone the recipient wants mail from.
- Open without subsequent negative action. A recipient opening the message and not marking it as spam is a mild positive signal. Opens alone don't build reputation fast, but their absence is a worse signal.
- Forward. Forwarding the message to another recipient is a strong signal that the content had value.
Manual warmup techniques that still matter
Automated warmup tools have appeared in the last year or two — Lemwarm bundled into Lemlist, a handful of other emerging products — and they work reasonably well for ongoing maintenance. But as of mid-2018, the tooling is still maturing, and every serious cold program mixes automated warmup with manual warmup techniques that the human team controls directly. Manual warmup is more operational work, but it produces stronger signals because the engagement is actually human.
Five manual warmup patterns worth knowing
- The colleague chain
- Send warmup emails to colleagues, ask for a short reply, maintain a small conversation thread. Real conversation in a normal business context is the purest form of positive signal. The downside is that it requires colleagues' time; it scales up to a team's reasonable bandwidth and no further.
- The customer loop
- If you have existing customers, they are your warmup asset. Ask a subset of customers if they'd engage with periodic updates from your team. Their replies build reputation while also producing legitimate customer communication value. Most customers are happy to participate if you frame the ask honestly.
- The personal network opt-in
- Personal contacts, friends, colleagues from previous jobs — people who know you personally. Ask them to reply when they get a warmup email. The relationship base makes the engagement organic.
- The internal calendar sync
- A simple trick: set up internal newsletters, meeting notes, or project updates that the team responds to in a normal business way. The domain's email sends and receives real work traffic, which builds reputation naturally as a byproduct of doing normal work.
- The warmup exchange network
- Several teams in similar situations agree to warm up each other's inboxes. You send to their network, they send to yours, both produce mutual engagement. Emerging warmup-as-a-service products like Lemwarm formalize this pattern, but the manual version works too and is free.
Manual warmup produces the best signals when it is spread across multiple recipients and conversations rather than concentrated on one. A domain warming up by sending exclusively to three colleagues produces a signal that reads "very small circle of internal correspondence" rather than "legitimate business with varied external relationships." Diversity of recipients matters as much as quantity of engagement.
What GDPR changed for cold in 2018
GDPR came into force in May of this year, and the ripples across cold email in Europe are still settling. The common (incorrect) reading is that GDPR made cold email illegal in Europe. It did not. But it did change the operational discipline required to send cold mail to EU recipients lawfully, and the teams that have not yet internalized that discipline are running higher legal risk than they realize.
What GDPR actually requires for B2B cold outreach
| Requirement | What it means for cold | Practical implementation |
|---|---|---|
| Legal basis | Most B2B cold relies on "legitimate interest" (Article 6(1)(f)) | Document your legitimate interest assessment before sending |
| Data minimization | Collect only what is necessary for the outreach purpose | Don't build elaborate enrichment profiles; keep the dataset narrow |
| Right to object | Recipients must be able to object easily | Clear opt-out in every message, honored promptly |
| Transparency | Recipient must be told how you got their data and what you plan to do with it | First-email transparency paragraph; linked privacy notice |
| Record keeping | You must be able to show compliance on request | Log every send, every opt-out, every legitimate-interest assessment |
| Data protection impact assessment | Not strictly required for most cold email, but useful documentation | Short internal document covering scope, purpose, safeguards |
The practical net effect on warmup is modest: your warmup content and your cold outreach content both need to satisfy GDPR requirements for EU recipients. This means shorter messages, clearer unsubscribe, and more careful data handling than was typical pre-May 2018. It does not mean you cannot warm up a domain for EU-focused cold outreach. It just means the operational hygiene around that outreach has to be tighter.
Non-EU recipients are unaffected by GDPR directly, but many cold programs now apply GDPR-style hygiene globally as a matter of operational simplicity. The additional effort is modest; the complexity reduction of having one policy rather than two is significant. This is the pragmatic default for most cold programs in mid-2018.
Measuring progress without kidding yourself
Warmup feels subjective in a way that most deliverability work does not. Are the engagement rates high enough? Is the reputation actually building? Can we ramp this week or should we hold? Without a measurement framework, these questions become guesses. With one, they become observable.
The metrics that tell you whether warmup is working
- Google Postmaster Tools domain reputation. Verify your warmup domain in GPT the day you register it. Watch the reputation trend. You want to see movement from "no data" or "Low" toward "Medium" during the warmup weeks. "High" is the destination but usually takes twelve weeks or more.
- Microsoft SNDS filter result. Enroll your sending IPs. "Green" is the target; "Yellow" during early warmup is acceptable; "Red" at any point means stop and investigate.
- Warmup-network engagement rates. Open rate 80%+, reply rate 30%+ across your warmup audience. If these are low, either your warmup content is weak or your warmup audience is too small for the sending volume.
- Seed list inbox placement. Run a small seed test at the end of each week. Send a warmup-style message to a handful of Gmail, Outlook, and Yahoo accounts you control. Where does it land? Inbox, promotions, spam? Track the distribution week over week.
- Bounce rate. Should be very low (under 1%) during warmup, because you're sending to audiences whose addresses you control. If bounces are appearing, your list has errors or your infrastructure has an issue.
- Authentication pass rate. SPF, DKIM, DMARC should all be passing at 99%+. Anything lower means there's a misconfiguration somewhere that needs fixing before cold outreach starts. For the broader framework of how to coordinate SPF, DKIM, and DMARC across multiple cold outreach domains, see authentication alignment and policy review for multi-domain senders.
When to slow down, when to push
The warmup curve is not a fixed schedule that every domain follows. It is a default cadence to adjust based on real signals. Knowing when to slow down, hold steady, or push faster is the skill that separates experienced operators from new ones.
| Signal | What it suggests | Action |
|---|---|---|
| GPT domain reputation drops or stalls | Something about recent sending is unhealthy | Hold volume; investigate content and engagement; do not ramp |
| Warmup engagement rates below 70% opens / 25% replies | Warmup network is too small or content is weak | Expand warmup network or diversify content; do not ramp volume yet |
| SNDS shifts from Green to Yellow | Complaint rate or filter placement degrading | Hold volume; review recent content for triggers; slow the ramp |
| Bounce rate above 2% | List hygiene issue | Pause new sends; clean the list; resume when bounces are back under 1% |
| Authentication pass rate drops below 99% | Configuration issue on one of the records | Fix immediately; do not ramp until authentication is stable |
| All metrics green for two consecutive weeks | Warmup is going well; room to move faster | Consider modestly faster ramp — 20% above default schedule |
| Seed tests land in inbox consistently at 90%+ | Strong reputation foundation | Standard ramp is appropriate; no acceleration needed |
| You receive a blocklist notification | Something triggered a manual report | Stop all sends immediately; investigate; delist; then assess whether to continue |
The temptation most teams face is to push when they see a green week, especially if leadership is asking when pipeline will arrive. Resist this unless the signals are consistently green for multiple weeks. A single green week is noise; two green weeks in a row is signal. Push after signal, not after noise.
After warmup: the maintenance discipline
Completing the warmup phase is not the end of the reputation work. The difference between a cold program that sustains deliverability over months and one that degrades after the initial six-week success is whether the team treats warmup as a permanent maintenance function or as a one-time setup project.
The ongoing discipline
Keep fifteen to twenty percent of each mailbox's daily volume dedicated to maintenance warmup traffic indefinitely. This is not wasted volume; it is the reputation-maintaining traffic that keeps your cold sends landing in the inbox. The specific destinations can vary — warmup network, internal correspondence, customer updates — but the proportion should stay roughly constant.
Treat any content change as a partial warmup reset. Switching from one subject line pattern to another, moving from short text to longer messages, introducing a new template — any of these changes the sending pattern the provider has learned to trust. Run the new pattern at lower volume for a few days before rolling it out broadly.
Treat any unexplained deliverability dip as a signal to pause. Pushing through a dip usually worsens it. Hold volume, investigate, understand the cause, fix it if fixable, and resume ramping only after the underlying signals have recovered.
Frequently asked questions
Can we shortcut warmup if we're using an established ESP with good IP reputation?
Partially. If you're sending through a shared IP pool with existing reputation, you skip the IP warmup portion. But your domain still has no reputation if it's new. Domain reputation and IP reputation are tracked separately by providers; both need to be built. The ESP's IPs help you; your domain still has to earn its own trust.
What if we don't have colleagues or customers for manual warmup?
Use one of the warmup-exchange networks emerging in the cold email tooling space. Lemwarm is the most developed option as of mid-2018. The automated networks are not as strong as real human engagement but they're adequate for getting started, and you can supplement with manual techniques as they become available.
Should we run warmup during the weekend?
Yes, but at lower volume. Providers observe sending patterns; a domain that sends only Monday-Friday looks different than one that sends some weekend traffic too. A small volume on Saturday and Sunday — five to ten emails per mailbox — keeps the pattern looking natural without producing odd engagement signals.
How do we know if our warmup is going well vs going poorly?
GPT domain reputation trending up, SNDS Green for primary IPs, warmup engagement rates at 80%+/30%+, bounce rate under 1%, authentication at 99%+. If all five are in place, warmup is working. If any are off, investigate before proceeding.
What happens if a warmup domain catches a blocklist?
Stop sending from it immediately. Investigate whether the cause is real (bad list quality, trap hit, compromised account) or spurious (opportunistic user report). Fix the root cause, delist from the blocklist, and decide whether to continue with the domain or retire it. A domain that hit Spamhaus SBL during warmup often doesn't recover well; it may be cheaper to start on a fresh domain.
Can we warm up multiple mailboxes in parallel?
Yes, and this is the pattern most serious programs use. Each mailbox warms up independently on its own timeline, and the program as a whole scales by adding mailboxes rather than by pushing individual mailboxes past their sustainable volume. Four mailboxes each sending 45/day deliver 180/day; you don't need one mailbox sending 180/day.
Does warmup content need to look like our eventual cold content?
Loosely yes, very similar no. Warmup content should be in the same broad style — business B2B tone, similar length range, similar link patterns — so the provider learns your normal pattern. Making it identical to your cold template is wasted effort; making it wildly different risks teaching the provider something unhelpful.
Closing perspective
The teams that run cold email programs successfully in 2018 have internalized a specific truth: warmup is not a phase that ends. It is an initial acceleration followed by a permanent maintenance discipline that keeps the sending surface healthy for as long as the program runs. The initial six-to-eight-week timeline is the startup cost. The fifteen-to-twenty-percent-of-volume maintenance cost is the ongoing rent. Teams that pay both consistently have cold programs that produce predictable pipeline. Teams that pay neither have cold programs that oscillate between brief success and long remediation.
The other pattern worth naming: the single biggest predictor of cold email success is not the content, the targeting, the sequencing tool, or the offer. It is whether the sending foundation is healthy. A warmed-up domain with good authentication, sensible daily caps, clean list hygiene, and maintenance warmup discipline can make a mediocre offer perform. A poorly warmed domain will make a brilliant offer invisible. Neither extreme is common in practice, but the direction of the gradient is consistent.
For teams building out new cold programs: internalize the timeline before making commitments to leadership. "Pipeline in four weeks" is not realistic from a cold start; "pipeline building from week five, meaningful volume by week eight, steady-state from week twelve" is. The more accurately expectations are set at the beginning, the less remediation work is needed mid-program when the early-week numbers don't match the projections made during setup.
For teams with existing cold programs running on warmed domains: audit the maintenance discipline. Is the warmup volume percentage actually being maintained? Are engagement rates on that warmup volume still healthy? Are the metrics that mattered during setup still being watched? The difference between a program that degrades silently over six months and one that stays healthy for two years is whether these disciplines persist after the initial launch excitement fades. Usually, they don't persist unless someone owns the specific responsibility. Assign the ownership; the rest follows. The broader planning framework for how reputation and deliverability evolve as programs grow — including how multiple domains fit into a rotation strategy — is covered in planning IP reputation and deliverability for expanding sending programs.