Email Warm-Up Explained: What It Does and How Long It Takes in 2026

A fresh sending domain has the same problem as a fresh résumé: nothing on it. Gmail, Outlook, and Yahoo have never seen this domain before, have no idea whether the messages it sends will be welcomed by recipients, and default to treating “unknown” as “probably spam.” Email warm-up is the slow, deliberate process of changing that — sending a small, gradually increasing volume of messages that get opened, replied to, and never marked as spam, so the receiving providers build a record that this domain is run by someone people actually want to hear from.

This sounds simple. Almost nobody does it correctly the first time. The mistakes that void weeks of warm-up are not technical — they are operational, and they show up most clearly in the third week, when a sender gets impatient and ramps too fast. This article walks through what warm-up is actually doing, how the typical six-week timeline breaks down, the mistakes that undo it, and how to tell when the domain is genuinely ready. It is the operational follow-up to the broader email deliverability guide — that one frames the three-layer model; this one drills into the reputation layer.

What warm-up is actually doing

When Gmail evaluates a message from a new sending domain, it has nothing to compare against. The provider’s spam model is built around per-domain reputation — a privately-held score that tracks how recipients have historically interacted with messages from that domain. A new domain has no history, which the model treats as a high-risk signal: messages from unknown domains get routed to spam or quarantine until the domain proves itself. Authentication setup (SPF, DKIM, DMARC — covered separately in our cold email authentication guide) is a prerequisite, but it is not reputation. Authentication says “this domain is allowed to send”; reputation says “this domain’s messages are welcome.”

Warm-up gives the domain a chance to prove itself in the safest possible way. The mechanics:

• Small volumes. Five to ten messages a day for the first week. Gmail won’t see this as a spam campaign — it sees it as normal correspondence.
• High engagement. Every message gets opened. About a third get replied to. None are marked as spam. The recipient signals are uniformly positive.
• Inbox recovery. Any message that lands in spam gets moved back to the inbox manually. Gmail records “user disagrees with our classification” — a strong reputational positive.
• Gradual scaling. Each week the volume rises, but the engagement stays roughly proportional. The domain looks like an actual person whose correspondence is growing naturally.

After six to eight weeks of this, the receiving providers have a track record showing that messages from this domain reliably get welcomed by recipients. The domain’s reputation score climbs into the range where new outbound messages default to the inbox rather than to quarantine. The investment up front buys you months of clean placement after.

The realistic six-week timeline

Real warm-up timelines vary slightly by provider and starting state, but the rhythm is consistent. We have warmed dozens of domains for client campaigns over the past two years, and the same shape repeats:

Week	Daily volume	Reply target	Notes
1	5–10	~33%	All messages to known mailboxes. Open every one.
2	10–25	~30%	Add a few external participants from a warm-up pool.
3	25–50	~25%	First “real” messages can start mixing in here.
4	50–100	~20%	If placement holds, begin small actual campaigns.
5	100–200	~15%	Cold campaigns become primary, warm-up secondary.
6	200–500	~10%	Domain is operational. Continue light warm-up indefinitely.

The reply rate target drops as volume rises because real cold campaigns rarely hit 30% reply rates — and a warm-up signal that doesn’t taper down looks artificial to the receiving filters.

The single most important rule across all six weeks: do not skip a step. If week three’s placement is below 70%, do not move to week four’s volume — repeat week three until placement stabilizes. Sending more messages on a weak reputation is the fastest way to make the reputation worse.

Mistakes that void warm-up

In our experience the same four mistakes show up over and over:

Ramping too fast. The most common failure. A sender warms for three weeks, sees decent placement, and immediately tries to send 1,000 messages on day 22. The volume jump itself looks suspicious to providers — a 10x increase in 24 hours is a hallmark of a compromised account. The domain gets throttled, and the previous three weeks of work are lost.

Mixing real outreach in too early. Real cold outreach is, on average, less engaging than warm-up traffic. Open rates are lower, reply rates are lower, deletions are higher. If real outreach starts in week one — instead of week three or four, when the domain has built a buffer — the average engagement drops below what providers expect from a “healthy” domain, and reputation stalls.

Inconsistent volume. Warm-up that sends 25 messages on Monday, zero on Tuesday, 60 on Wednesday confuses the rate of reputation accumulation. Providers prefer to see steady, predictable patterns — they are inferring intent from behavior, and erratic behavior reads as suspicious. Send roughly the same volume every business day.

Treating warm-up as a one-time setup. Warm-up does not end at week six. A small ongoing volume of warm-up traffic — even 10% of total send — keeps engagement signals positive in periods when real campaigns underperform. The senders who never see deliverability drops are the ones who maintain warm-up indefinitely. The senders who hit “fully warmed” and stop warm-up altogether tend to lose placement within a quarter.

Recovering after a reputation hit

Warm-up isn’t only for new domains. Established domains that take a reputation hit — a spam complaint surge, a list with high bounces, a sudden 10x volume spike — need to be re-warmed back to clean placement. The protocol:

1. Stop all real campaigns immediately. Continuing to send to a damaged domain compounds the problem.
2. Drop volume by 80% for two weeks. Resume warm-up at week-two volumes (10–25 messages a day) on the warm-up pool only.
3. Verify the underlying cause is fixed. Whatever caused the hit — bad list, broken authentication, unsubscribed recipients still receiving — must be resolved before resuming.
4. Ramp slowly back. Treat the domain as if it were brand new from this point. Two to three weeks of clean warm-up, then small real campaigns, then full volume.

Re-warming typically takes three to four weeks. Senders who try to rush it back into full campaigns end up burning the domain entirely — at that point, the only fix is to register a new sending domain and warm it from scratch.

Tools and their real tradeoffs

Warm-up tools route messages between participating mailboxes and simulate engagement. They are all variations on the same idea, and they differ less in capability than in pool quality and integration depth.

Mailwarm. One of the originals. Standalone service, $69/month per mailbox. Pool quality is decent but smaller than newer entrants. Good if you want a tool that does only warm-up and nothing else.

Lemwarm. Lemlist’s warm-up product. Integrates directly with Lemlist campaigns, so the same domain warming + sending happens in one workflow. About $29/month per mailbox standalone, included in some Lemlist plans. Pool is large and active.

Warmup Inbox. Aggressive pricing ($19/month for the first mailbox). Pool size is reasonable. Has a free tier limited to a tiny daily volume.

Instantly built-in warm-up. Bundled with Instantly Sender, no additional fee. Pool is one of the larger ones in the industry because of Instantly’s user base. If you are already sending campaigns with Instantly, no reason to look elsewhere.

Smartlead built-in warm-up. Same logic as Instantly’s offering. Pool is solid. Bundled with the platform.

What none of these will tell you: the pool you participate in matters more than the tool. A small or inactive pool means your messages bounce around the same 200 inboxes, which providers can detect as artificial behavior. Larger, more diverse pools look more natural. Of the options above, Instantly and Smartlead currently have the largest pools because they have the largest user bases.

For agencies running warm-up across many domains, the difference is operational. A tool that handles 50 domains in one dashboard, with automatic reply-rate tuning, saves more time than the cheapest standalone option.

How to know warm-up is actually working

The metric you care about is inbox placement rate, not anything the warm-up tool reports. A warm-up tool will tell you “your domain is 78% warmed” — that number is the tool’s internal score, not a real placement measurement. The only ways to actually verify:

• Seed testing weekly. Send a test message to a panel of 20–30 inboxes spread across Gmail, Outlook, Yahoo, and corporate Microsoft 365 domains. Check where each one lands. Tools like GlockApps run this automatically for $30–60 per test.
• Google Postmaster Tools. Domain reputation rating climbs from “No data” through “Bad” → “Low” → “Medium” → “High” as warm-up progresses. “Medium” means you can send cold campaigns. “High” means you have headroom for volume.
• A small real campaign. After week three or four, send a real campaign of 30–50 messages to a curated list. If reply rates and bounces look reasonable, warm-up is on track. If bounce rate exceeds 5% or placement is obviously bad, something is wrong with the warm-up — investigate before scaling.

A domain that hits “High” on Postmaster Tools, places at 80%+ in seed tests, and survives a 50-message real campaign with no spam complaints is genuinely warmed. Anything short of that is the warm-up tool’s marketing department talking, not the receiving providers.

The discipline to slow down after a hit — and the discipline to keep warm-up running indefinitely instead of “checking it off the list” — is what separates senders whose domains last years from senders who burn through one domain every quarter.