How many inboxes do you actually need for cold email?
The sending math we use to size domains and inboxes for a programme, and why the answer is almost always more inboxes sending less.
The first question on most kickoff calls is some version of “how many emails can we send?” The better question is how many inboxes you need to send that volume without burning a domain. Here is the math we use on every programme.
The rule: low volume per inbox, many inboxes
Mailbox providers judge each sending inbox on its own behaviour. An inbox that sends a few dozen plain-text emails a day and gets replies looks like a person. One that sends hundreds of near-identical emails looks like what it is. So we keep each warmed inbox to a modest daily send and scale by adding inboxes, never by pushing one inbox harder.
Our working numbers:
- Daily sends per warmed inbox: around 15 to 30, depending on the provider and how the domain is ageing.
- Inboxes per domain: two or three. More than that and one bad day hurts too many inboxes at once.
- Warm-up: about two weeks before the first cold send, and it keeps running in the background after launch.
Working it out
Start from contacts, not emails. If you want to reach 6,000 new contacts a month with a three-step sequence, that is roughly 18,000 emails a month, or about 820 per working day across 22 days.
| Input | Value |
|---|---|
| New contacts per month | 6,000 |
| Steps per sequence | 3 |
| Emails per working day | ≈ 820 |
| Sends per inbox per day | 25 |
| Inboxes needed | ≈ 33 |
| Domains at 3 inboxes each | 11 |
We then add a reserve of around 20% so that any domain dropping below our placement threshold can be rested and swapped out without slowing the programme. That reserve is the difference between a campaign that survives a bad week and one that stops.
Why not just send more from fewer inboxes?
Because the cost is not linear. A single inbox pushed past its comfortable volume starts landing in spam, and spam placement drags reply rates down across every inbox on that domain. A few extra inboxes cost very little each month; a burned domain costs you weeks of warm-up and the pipeline you would have booked in that time.
The checks that keep it healthy
- Seed tests twice a week across Google and Microsoft inboxes.
- Any domain under 85% inbox placement is pulled, rested and replaced from the reserve.
- Bounces held under 1.5% by verifying every contact before it reaches a sequencer.
- Open tracking off, because tracking pixels hurt placement and open rates are unreliable anyway.
None of this is exotic. It is just discipline, applied every week, which is exactly why it is worth having someone own it.