"How many mailboxes per domain" gets asked as if there's a universal constant, but the honest answer is a function of three inputs: your daily send cap per mailbox, your target monthly volume, and how much risk you're willing to concentrate on any single domain. Here is the actual math, not a rule of thumb.
Start with the daily cap per mailbox
A fully warmed mailbox on a healthy domain can typically sustain 150 to 200 outbound sends a day without damaging engagement metrics. A mailbox still in its first six weeks should be held far lower (see the warmup ramp for specifics). This cap is per mailbox, not per domain, and it's the number everything else is derived from.
How many mailboxes belong on one domain
The constraint here isn't technical. DNS doesn't care whether a domain has 2 mailboxes or 50. The constraint is blast radius: if a domain's reputation degrades, every mailbox on it degrades with it, and every open conversation and pending reply on that domain becomes collateral damage. Most fleets land between 3 and 10 mailboxes per domain, with the lower end reserved for higher-stakes or higher-volume sending and the upper end for lower-risk, well-established programs.
Monthly send volumeUnder 5,000
- Mailboxes / domain
- 2-3
- Reasoning
- Small blast radius matters more than domain-count overhead
Monthly send volume5,000-25,000
- Mailboxes / domain
- 3-5
- Reasoning
- Balance of manageability and domain administration cost
Monthly send volume25,000-100,000
- Mailboxes / domain
- 5-8
- Reasoning
- Domain count is growing regardless; keep per-domain risk moderate
Monthly send volume100,000+
- Mailboxes / domain
- 8-10
- Reasoning
- Established programs with steady-state reputation and monitoring in place
| Monthly send volume | Mailboxes / domain | Reasoning |
|---|---|---|
| Under 5,000 | 2-3 | Small blast radius matters more than domain-count overhead |
| 5,000-25,000 | 3-5 | Balance of manageability and domain administration cost |
| 25,000-100,000 | 5-8 | Domain count is growing regardless; keep per-domain risk moderate |
| 100,000+ | 8-10 | Established programs with steady-state reputation and monitoring in place |
Working backward from a monthly target
Once you know the per-mailbox cap and your mailboxes-per-domain ratio, the domain count falls out of a simple formula: divide your target monthly sends by the product of mailboxes per domain, sends per mailbox per day, and working days in the month.
Fleet sizing formula
domains_needed = target_sends_per_month
/ (mailboxes_per_domain * sends_per_mailbox_per_day * sending_days_per_month)
// Worked example:
// target = 60,000 sends/month, mailboxes_per_domain = 5,
// sends_per_mailbox_per_day = 150, sending_days_per_month = 20
domains_needed = 60000 / (5 * 150 * 20) = 60000 / 15000 = 4Blast radius: why concentration is the real risk
- A blocklist listing or a DMARC enforcement mistake takes out every mailbox on the affected domain simultaneously, not just the one that triggered it.
- Concentrating volume onto fewer domains makes each domain's sending pattern look more aggressive to inbox providers evaluating volume-per-domain trends.
- Recovery is per-domain, not per-mailbox. A damaged domain needs its own re-warm cycle even after the root cause is fixed, which is slower than swapping one bad mailbox.
- Client or campaign isolation gets harder to reason about once too many unrelated conversations share one domain's reputation.
Worked example: 50,000 sends a month
At 50,000 sends a month with a moderate 4 mailboxes per domain and a conservative 120 sends per mailbox per day over 20 sending days, the math calls for roughly 5.2 domains. Round up to 6 to leave headroom for a domain in re-warm or under review at any given time. That headroom isn't waste; it's the buffer that keeps one flagged domain from forcing you to pause outbound entirely.
Always size in a reserve
Plan for 15 to 20% more domain capacity than the math strictly requires. Domains occasionally need to be pulled for review, and a fleet with zero slack turns a single incident into a full outbound stoppage.