Everyone asking this question wants a single safe number. There isn’t one. The daily volume an inbox can handle depends on how long it has been sending, what its metrics looked like last week, and how the rest of your infrastructure is behaving. Treating volume as a fixed setting is how teams burn domains.
The useful mental model is a ramp per inbox, multiplied across enough inboxes to hit your monthly target, with brakes that engage the moment signals dip. This guide covers the per-inbox norms practitioners actually use, why they exist, and two worked sending plans so you can see what the numbers look like at small and mid scale.
The honest answer is a ramp, not a number
Mailbox providers score senders on behavior over time, not on any single day. That score, your sender reputation, is built from bounce rates, complaint rates, engagement, and how natural your sending pattern looks. A warmed inbox that has sent 25 emails a day for three months with clean metrics is a known quantity. A week-old inbox sending the same 25 looks like the opening move of a spam operation, because that is exactly the pattern spam operations open with.
So the question is never “how many emails can I send per day.” It is “how many can this specific inbox send today, given its history.” Early on the answer is tiny. Months in, it is comfortable but still capped. The number moves; the discipline of moving it slowly is the whole game.
Per-inbox daily norms at steady state
Once an inbox is fully warmed, most practitioners cap it somewhere between 20 and 50 cold emails per day, and the teams most protective of deliverability usually settle around 25 to 30. Two reasons the ceiling sits there:
- Providers watch per-account behavior. A normal salesperson does not fire 200 near-identical emails to strangers daily. Staying inside the volume range of a busy human is the simplest way to avoid pattern-based filtering.
- Headroom is insurance. An inbox running at half of what it could theoretically handle can absorb a bad list segment or a complaint spike without tipping into the spam folder. An inbox running at its limit cannot.
Two details people miss. First, warmup traffic typically keeps running in the background even after ramp, so the inbox’s total daily activity is cold sends plus warmup sends. Second, the cap is for sends spread across the working day. Twenty-five emails dripped between 8am and 5pm reads human; the same 25 fired in one burst at 9:00 does not.
Why fresh domains start tiny
Cold outreach should never run on your primary domain; it runs on dedicated sending domains bought for the purpose, so a reputation problem never touches your company’s real email. The trade is that those domains start with zero history.
A new domain and its inboxes go through warmup before any cold volume: low-level traffic that draws opens and replies so the first behavior providers record from the domain is positive. Practitioners typically give this 2 to 4 weeks. After that, cold sending starts at around 5 to 10 emails per inbox per day, then steps up gradually. Increases of roughly 5 to 10 per inbox per week are a common pace. Doubling overnight is the classic way to undo a month of warmup in an afternoon.
No setting in any tool changes this timeline. Products that promise instant volume on fresh domains are spending reputation the domain does not have yet.
Scale with more inboxes, not a louder one
When you need more volume, the answer is horizontal: more inboxes and more domains, each kept inside the same modest per-inbox cap. The norm most teams use is 2 to 3 inboxes per domain, so a domain rarely sends more than 50 to 90 cold emails a day in total.
This buys you two things. Capacity, obviously: 6 inboxes at 25 a day is 150 daily sends without any single account looking unusual. But more importantly, blast radius control. If one domain’s metrics slip, you rest it and shift its volume to healthy domains while it recovers. If all your volume runs through one overworked domain, a reputation dip takes your entire program offline at once.
The corollary: budget for domains and inboxes as a real line item from day one. Meaningful volume on one domain is not a cost saving, it is a single point of failure.
How warmup and throttling interact
Warmup and throttling are the two controls on the same dial. Warmup builds the positive history; throttling keeps the sending shape human while it does. In practice that means:
- Daily caps per inbox that follow the ramp schedule, not your pipeline goals.
- Randomized gaps between individual sends, minutes apart, not a fixed metronome.
- Sending windows aligned to business hours in the recipient’s region.
- Warmup traffic continuing alongside cold sends so engagement signals stay mixed and positive.
When these are set correctly, volume increases feel painfully slow. That is the point. Every shortcut shows up two weeks later as a placement problem you then spend a month fixing.
The signals that say slow down
A ramp only works if it can also go backwards. Watch these per inbox and per domain, weekly at minimum:
- Bounce rate. Keep it under 2 to 3 percent. Above that, stop sending to the offending segment and fix the list before resuming. A bounce spike on a young, low-volume inbox is especially damaging because there is so little good history to dilute it.
- Spam complaints. Google’s published bulk sender guidance tells senders to stay under a 0.3 percent spam complaint rate and to support one-click unsubscribe. Treat 0.3 percent as the cliff edge, not the target; practitioners aim well below it.
- Reply rate trend. A sudden drop in replies across an otherwise unchanged campaign often means placement degraded before any tool flagged it.
- Placement tests. Periodic inbox placement checks show where your mail actually lands, which is not the same as what your sending tool reports as delivered.
When any of these dip, cut that domain’s volume immediately, usually by half or more, and let it recover. Holding your daily number steady through a dip is choosing the monthly target over the domain, and the domain always loses.
Worked example: 3,000 emails per month
Call it roughly 150 sends per business day across about 20 sending days. A typical build:
- Infrastructure: 3 dedicated domains, 2 inboxes each, 6 inboxes total. Steady state is 25 per inbox per day.
- Weeks 1 to 2: warmup only. Zero cold sends.
- Week 3: 5 to 10 cold sends per inbox, roughly 30 to 60 per day total.
- Week 4: 10 to 15 per inbox, roughly 60 to 90 per day.
- Weeks 5 to 6: step to 20 to 25 per inbox and hold at about 150 per day.
Full target volume arrives about 6 weeks after domain setup. The first month delivers well under 3,000, and a plan that assumes otherwise is a plan to skip the ramp.
Worked example: 15,000 emails per month
Now you need roughly 750 sends per business day, which at 25 per inbox means about 30 inboxes across 10 to 15 domains. The numbers scale; the method changes in one important way: you ramp in cohorts.
- Cohort structure: bring domains online in groups of 3 to 5 (6 to 10 inboxes), starting a new cohort every 1 to 2 weeks rather than warming everything simultaneously.
- Each cohort follows the same ramp as the small example: 2 weeks of warmup, then stepped increases over 3 to 4 weeks.
- Timeline: the first cohort is at full volume around week 6, and the full 750 per day typically lands somewhere in weeks 8 to 12 depending on how cleanly each cohort’s metrics hold.
- Build spare capacity. Provision 10 to 20 percent more inboxes than the math requires, so you can rest a wobbly domain without missing the monthly number.
At this scale the operational load is real: dozens of inboxes to monitor, ramp schedules per cohort, and rotation decisions every week. The sending itself is the easy part.
If you would rather not manage the ramp
Everything above is learnable, and if you run outbound in-house this is the discipline to build. It is also the unglamorous work clients hire us to make disappear. Sendful’s managed system, The Outbound Engine, stands up the dedicated domains and inboxes, runs the warmup and the ramp schedules, and makes the rest-and-rotate calls per domain so you never have to track per-inbox caps yourself. Infrastructure is typically warming and live within 2 weeks of kickoff, we never send from your primary domain, you own the domains, lists, copy, and data, and you get weekly reporting with positive replies routed straight to you. If you want a second opinion on your own ramp math, book a call. You will leave with a custom outbound plan for your volume target, whether or not we work together.