How Long Should You Warm Up a New Email Account?

The honest answer most people don't want: a brand-new mailbox on a brand-new domain needs roughly three to four weeks of warm-up before it can carry real cold volume, and you should add another one to two weeks after registering the domain before you send anything at all. Call it five to six weeks from card swipe to a mailbox doing 25-30 cold emails a day.

That number annoys founders who bought a domain on Monday and want a campaign live on Wednesday. It also annoys people who've read that warm-up is a scam and you should just start sending. Both camps are reacting to the same real problem: warm-up is sold as a fixed calendar when it's actually a reputation-building process with wildly different timelines depending on where you're starting from.

{Here's how to think about it.}

The short answer by scenario

Those ranges assume you're targeting normal cold-outreach volume — somewhere in the 20-40 emails per mailbox per day band we cover in how many cold emails you can send per mailbox per day. If your target is 10 a day for a founder-led ABM motion, you can shorten every row. If you're planning 50+, lengthen them and expect trouble anyway.

For the day-by-day volume ladder, see how to warm up an email mailbox for cold outreach — no point repeating the schedule here. This post is about the duration question underneath it.

Why the calendar isn't the real answer

Mailbox providers don't have a timer running on your domain. Gmail and Microsoft are building a statistical profile: how much you send, how consistently, to whom, how those recipients react, and whether your authentication holds up. Time only matters because it's the axis all that evidence accumulates along.

Which means two mailboxes that both "warmed up for 30 days" can be in completely different shape. One sent 400 total messages that got opened, replied to, and dragged out of Promotions. The other sent 400 messages into a reciprocal warm-up pool where bots clicked "not spam" on each other's junk. The first has a reputation. The second has a log file.

So the useful question isn't "how many days" — it's "have I generated enough real positive engagement signal, at a consistent enough cadence, that a volume increase looks like growth instead of a spike?"

The five things that actually set your timeline

• Domain age. A domain registered last week is the single strongest negative signal you can carry. Filters have seen too many throwaway domains blast for 72 hours and disappear. Two weeks of DNS sitting there with valid SPF, DKIM, and DMARC before your first send is cheap insurance.
• Provider. Google Workspace tends to build trust faster and gives you Postmaster Tools to watch it happen. Microsoft 365 is slower and more opaque — expect to add about a week, and expect placement to Outlook-hosted recipients to lag Gmail placement even when everything looks fine.
• Target volume. Warm-up isn't building an absolute reputation, it's building the credibility to sustain a specific level. Going to 15/day takes far less runway than going to 40/day. Decide your ceiling first, then work backwards.
• List quality. A warm-up can't survive the first real campaign if that campaign hits 12% invalid addresses. Bounce rate is a reputation event, not a delivery inconvenience. Verify before you ramp, not after.
• What you're sending. A plain-text, one-link, no-image message from a warmed mailbox behaves very differently from a four-image HTML template with a tracking pixel and a shortened link. Content risk and reputation risk compound. Run drafts through spam scoring before you decide the mailbox is the problem.

Readiness signals that beat the calendar

Before you push a mailbox from warm-up into live campaigns, check these. If they're all green at day 18, go. If they're not green at day 35, don't.

• Inbox placement. Seed tests landing in the primary inbox for at least 90% of Gmail seeds and 80% of Microsoft seeds. Promotions-tab placement counts as a warning, not a pass.
• Zero authentication failures. SPF, DKIM, and DMARC aligned on every message, verified from actual received headers — not just from a DNS lookup tool.
• Reply behavior. Your warm-up messages are getting genuine replies inside the workspace, not just opens. Opens are the weakest signal in the stack and the easiest to fake.
• Bounce rate under 2% on whatever real sending you've done so far.
• Zero spam complaints across the entire warm-up window. One complaint in 300 sends is a 0.33% rate — already above where Gmail starts getting unhappy.
• Consistent daily volume for the last 7 days. Gaps and spikes read as bot behavior. A mailbox that sent 0, 0, 40, 0, 35 is worse off than one that sent 12 every day.

How long is too long

There's a diminishing-returns point, and it arrives sooner than the warm-up-tool industry implies. Past about four weeks of pure warm-up with no real sending, you're mostly paying for idle infrastructure. The reputation gains flatten; what moves the needle after that is actual cold sending that generates actual replies.

The bigger cost is opportunity. Six mailboxes sitting in warm-up for 90 days is a quarter of pipeline you didn't build. If you're stalling because you're nervous, the fix is to start smaller — 5 real cold emails a day into your best-fit, most-likely-to-reply segment — rather than to warm-up for another month.

Warm-up doesn't end, it just gets quieter

The most common mistake I see isn't warming too briefly. It's treating warm-up as a phase that finishes. A mailbox running cold campaigns at 30/day with a 2% reply rate is generating far less positive signal per message than it was during warm-up, and it's generating negative signal — deletes without reads, occasional complaints — for the first time.

Keep a low-level warm-up stream running underneath live campaigns indefinitely. Five to ten real, labeled, replied-to messages a day is enough to keep a baseline of positive engagement in the mix. It also gives you a canary: when warm-up placement starts slipping, you know before your campaigns do.

And re-warm after any gap. Two weeks off is fine. Two months off means the provider has largely forgotten you — restart at a third of your old volume and rebuild over a week or so.

Mistakes that turn 3 weeks into 3 months

• Jumping from 8 emails/day to 40 the day warm-up "ends." Step up 20-30% every few days instead.
• Warming up on the domain you use for billing, support, and customer email. Buy a separate sending domain.
• Using reciprocal warm-up networks, where the same few thousand mailboxes trade artificial engagement. Providers have been pattern-matching this for years.
• Sending the first real campaign to your oldest, coldest, least-verified list. Send it to your best segment.
• Ignoring Microsoft. Testing only Gmail placement and being blindsided when 40% of your list is Outlook-hosted.
• Running warm-up and campaigns from the same mailbox with no shared volume ceiling, so the mailbox quietly does 60/day total.

How Sendvanta handles the ramp

Sendvanta warms mailboxes with real, labeled messages inside your own workspace — no shared pool, no proxied IPs — and layers a health score across mailbox, domain, DKIM, and IP that automatically slows or pauses sending when risk thresholds are crossed, which is usually the thing that saves a mailbox from a ramp that went too fast. Drafts get SpamAssassin-scored before send, seed tests measure actual inbox placement, and engagement classification separates verified replies and visits from security-scanner clicks so your readiness signals aren't inflated. You can connect your own Gmail, Microsoft 365, or SMTP mailboxes free on every plan, including the $0 free tier (1,000 active leads, 3,000 emails/month, up to 10 mailboxes) — enough to warm a full set of mailboxes before you pay anything. Paid plans start at $29/mo, and if you'd rather not provision infrastructure yourself, managed mailboxes are $4/mo each with a one-time $20 domain registration. More at features.