Most people pick an inbox count by guessing, then find out in month two that it was wrong in one direction or the other.
The calculation takes thirty seconds:
Inboxes = monthly sends ÷ (daily limit per inbox × sending days)
The only input people get wrong is the daily limit, and it changes the answer by a third.
The numbers
22 sending days a month. Weekdays only, because weekend sending into a B2B inbox produces almost nothing and still costs domain reputation.
| Monthly sends | At 15/day | At 25/day | At 30/day |
|---|---|---|---|
| 50,000 | 152 | 91 | 76 |
| 100,000 | 303 | 182 | 152 |
| 170,000 | 515 | 310 | 258 |
| 300,000 | 909 | 546 | 455 |
Look at the 170,000 row. The difference between 15 a day and 25 a day is 205 inboxes for identical output. Inboxes are a recurring cost, so that gap is usually the largest controllable line in an outbound budget.
Why you cannot just start at 25
A fresh domain has no reputation, and volume that climbs too fast looks like a spike. A spike looks like a spammer.
| Stage | Per inbox, per day |
|---|---|
| Fresh out of warmup | 10 to 15 |
| Weeks 2 to 4 | 15 to 20 |
| Established, bounce under 2% | 25 |
| Mature fleet, clean history | 30 |
| Above 40 | risk climbs sharply |
So the honest answer to "how many inboxes" is two answers. The number you need in month one, and the number you need once the fleet has settled. Budget for the first, plan for the second.
Three inboxes per domain, not more
The domain is the unit that gets blacklisted, not the inbox. Ten inboxes on one domain means a bad week removes ten sending seats at once. Three inboxes across many domains means it removes three.
So 300 inboxes is about 100 domains. That ratio is not a preference, it is how you distribute risk.
What it costs
| Item | Typical |
|---|---|
| Domain | $10 to $15 a year |
| Inbox via an infrastructure provider | $1.50 to $3 a month |
| Inbox bought direct from Google or Microsoft | $6 to $7.20 a month |
On a 300-inbox fleet that difference is over $1,000 a month for the same mailboxes. It is the first thing worth checking on any setup that feels expensive.
The mistake underneath all of it
People buy inboxes to hit a sending target, then find the list cannot feed them. Sending capacity is easy to buy. Contacts are finite.
Before buying anything, divide planned monthly sends by three touches. That is how many unique people you need every month. If your market does not contain that many, the inbox count is the wrong problem, and I wrote about that in why your TAM is smaller than your inbox count.
If emails are already landing in spam, fleet size is not the issue either. Work through the order to check things in first.