Email throttling is deliberately controlling the rate and volume of your sends to match what mailbox providers consider natural. Two layers: how many emails a mailbox sends per day, and how those sends are spaced across the day. Get both wrong and your sending looks like a machine, which is what fil
Email throttling is deliberately controlling the rate and volume of your sends to match what mailbox providers consider natural. Two layers: how many emails a mailbox sends per day, and how those sends are spaced across the day. Get both wrong and your sending looks like a machine, which is what filters are built to catch.
The working numbers for cold email: 25 to 50 per mailbox per day at an established account, spread across business hours with randomized gaps, and ramped by no more than 25% week over week. New domains start at 20 to 30 a day. This post covers the volume numbers, the intra-day pacing, the ramp schedule, and the signals that tell you to stop.
Email throttling is the practice of pacing outbound email so its rate and volume resemble normal human sending rather than a burst. It operates at two levels: a daily volume cap per sending mailbox, and spacing between individual sends across working hours. Throttling protects sender reputation by keeping any mailbox or domain from spiking, which mailbox providers treat as a spam signal.
Controlling two things. The daily total per mailbox, so no account sends more than its reputation can support. And the pattern within the day, so 40 emails go out over eight hours with irregular gaps rather than in a two-minute burst at 9am.
Your sending platform does the mechanical part once you set the parameters: a per-mailbox daily cap, a sending window, and randomized intervals. Throttling is the difference between a mailbox that looks like a person working through a list and one that looks like a script. It is not about sending less over time. It is about sending the same volume in a shape that does not trip a filter.
Three reasons providers punish un-throttled sending.
Bursts look automated. A human does not send 50 identical-structure emails in 90 seconds. Providers rate-limit directly: Gmail's rolling recipient window and Outlook's 30 messages per minute are hard ceilings a sequencer must stay under. And a volume spike relative to your history, going from 40 a day to 400 overnight, reads as a compromised account or a spam run and triggers scrutiny.
Throttling keeps your footprint smooth and small at every level. See why are my emails going to spam.
Per mailbox, for cold outreach:
| Situation | Per mailbox / day |
|---|---|
| New domain, week 1 | 20 to 30 |
| New Microsoft 365 tenant, first 30 days | under 30 |
| Established account, conservative | 10 to 20 |
| Established account, standard | 25 to 50 |
| Pre-warmed Microsoft 365, from day one | 100 to 150 |
For Google Workspace and Microsoft 365, keep cold outreach under 100 per mailbox per day even at fully established accounts, because cold mail generates lower engagement and higher complaint rates than opt-in marketing. See how many cold emails to send per day and how many email accounts per domain.
The daily cap is half the job. The other half is spreading those sends out.
Scaling per-mailbox volume follows one rule: increase by no more than 25% week over week, and plateau at any sign of trouble.
This runs alongside domain warm-up, which covers the first few weeks of a new domain's life at the low end of these numbers.
Say you want to reach 6,000 cold emails a month, roughly 300 a day on weekdays, from a fresh set of domains.
If week 2 had shown rising deferrals, you would have held at 25 a day for a week, diagnosed, and only stepped up once the deferrals cleared. The ramp is a ceiling on ambition, not a schedule you owe the calendar. See how many email accounts per domain.
Stop raising volume, and consider lowering it, when you see any of these:
Any one of these means your current volume is at or past what your reputation supports. Hold, diagnose, and only resume ramping once the signal is clean. See email domain reputation check.
At scale, throttling is a fleet-level job. You have a total daily target, and you distribute it across many mailboxes on many domains so each mailbox stays in its 25-to-50 band and each domain stays under its per-domain ceiling.
Inbox rotation is the mechanism: the platform assigns each recipient to a mailbox in turn, paces each one, and keeps the whole pool smooth. Split the fleet across Google and Microsoft so provider-specific rate limits never bottleneck the whole operation. See best ESP for cold email.
The 25%-a-Week Ramp: raise per-mailbox volume by at most 25% week over week, and plateau the moment 4xx errors, complaints, or placement move the wrong way. Speed of scaling is capped by the slowest health signal.
The rule keeps scaling honest. It is tempting to double a mailbox's volume the week a campaign is working, and that is exactly when a domain gets flagged, because reputation lags volume by days. A 25% weekly step is slow enough that a problem shows up in your metrics before it becomes a block, giving you room to plateau.
The plateau condition is the important half. Ramping is conditional on every health signal staying clean: 4xx errors flat, spam rate under 0.1%, bounces under 2%, replies holding, placement steady. If any one degrades, you stop raising volume regardless of how good the others look, because the weakest signal is the one that predicts a block. The quotable version: "Scale at the speed of your worst metric, not your best campaign."
Resume the ramp only after the degraded signal recovers.
Throttling controls the rate you send. It does nothing about what you send to. The health signals that force a plateau, bounces and complaints, are driven by list quality, so a bad list caps your throttling ramp low no matter how carefully you pace.
InboundLabs is a sales intelligence platform built on a database of 280M verified B2B contacts with 98% email deliverability on verified contacts. It layers buyer intent signals on firmographic data and lets you filter by industry, headcount, region, and title, so your bounce and complaint rates stay low and your ramp can keep moving. Monthly plans, no annual lock-in. Free to start, no credit card required.
See how InboundLabs finds verified contacts instantly → inboundlabs.app
Email throttling paces your sends to look human at two levels: a daily cap per mailbox, and spacing across business hours with randomized gaps. For cold email, run 25 to 50 per mailbox per day at established accounts, 20 to 30 on new domains, and stay under 100 even when fully warmed. Ramp per-mailbox volume by no more than 25% a week, and plateau the instant 4xx errors, complaints, or inbox placement move against you. At scale, distribute the daily total across a rotated fleet so no mailbox or domain ever spikes.
Deliberately pacing your outbound email so its rate and volume look like normal human sending rather than a burst. It works at two levels: a daily volume cap per sending mailbox, and spacing between individual sends across working hours. It protects sender reputation by keeping any mailbox or domain from spiking.
25 to 50 at an established account, 10 to 20 if you want to be conservative, and 20 to 30 on a new domain in its first week. Keep it under 100 even when fully warmed, because cold email generates more complaints and lower engagement than opt-in mail. Scale by adding mailboxes, not raising this number.
By no more than 25% week over week, and only while every health metric stays clean: 4xx errors flat, spam rate under 0.1%, bounces under 2%, reply rate holding, inbox placement steady. Plateau the moment any one degrades. Reputation lags volume by days, so a slow ramp catches problems before they become blocks.
Sending too fast or in bursts, exceeding a rate limit like Outlook's 30 messages per minute, or spiking volume relative to your history. The provider responds with 4xx deferrals or a temporary send block. The fix is slower, more even pacing, not a request to raise the limit.
Generally no, or at much lower volume. A person working through a prospect list sends during business hours on weekdays. Sending your normal weekday volume on a Saturday is a pattern that does not match human behavior and adds nothing, since reply rates on weekend cold email are low anyway.
Use inbox rotation. Set a total daily target, and let your sending platform distribute it across every mailbox on every domain, pacing each one so it stays in its 25-to-50 band and each domain stays under its per-domain ceiling. Split the fleet across Google and Microsoft so no single provider's rate limit bottlenecks everything.
LSI keywords: send rate pacing, daily volume cap, sending window, randomized intervals, ramp schedule, 4xx errors, deferrals, inbox rotation, per-mailbox limit, warm-up, rate limit, sender reputation
The List-Unsubscribe header gives a mail client a machine-readable way to opt a recipient out, and since 2024 the one-click version defined by RFC 8058 is mandatory for bulk senders to Gmail and Yahoo, meaning anyone sending 5,000 or more messages a day to personal accounts at those domains.
Yes, open tracking hurts deliverability, mildly, and mostly through the tracking domain rather than the pixel itself. If the domain hosting your tracking pixel has a poor reputation from other senders sharing it, your email can inherit that spam classification. Practitioners report that disabling op
The best email warm up tool depends on one number: how many mailboxes you run. At 3 mailboxes, per-inbox pricing is fine and any tool works.
The best setup for cold email is not a traditional email service provider. It is real Google Workspace or Microsoft 365 mailboxes, on secondary domains, connected to a sending platform like Instantly, Smartlead, or lemlist.
No commitment. No credit card. Just 50 free verified contact lookups.