Most outbound setups are sized by gut feel. Someone buys ten domains because ten sounded reasonable, adds two mailboxes each, and then discovers in week three that the campaign needs three times the capacity, or that half the mailboxes are sitting idle while the other half are being pushed past what is safe.
There is a better order of operations. Start with the outcome you actually care about, work backwards through your funnel to a daily send number, and only then decide how many mailboxes and domains you need. The math is simple. The discipline is in using conservative inputs and leaving room for the things that always go wrong.
Start From the Meeting Goal, Not the Send Count
Send volume is an input, not a goal. Nobody gets paid for emails sent. Pick the number that matters to the business, usually booked meetings per week, and treat everything else as a consequence of it.
To work backwards you need two conversion rates from your own history:
- Positive reply rate: the share of prospects contacted who reply with interest over the full sequence.
- Booking rate: the share of positive replies that turn into a held meeting.
If you have no history, do not borrow a flattering number from a vendor's landing page. Use deliberately cautious placeholders, run for two or three weeks, then replace them with real data. The examples below use a 1% positive reply rate and a 50% booking rate. Those are planning assumptions chosen to be cautious, not benchmarks, and your numbers could be meaningfully higher or lower depending on offer, list quality and market.
Turn a Meeting Goal Into Daily Sends
The chain runs in four steps: meetings to positive replies, positive replies to prospects, prospects to total emails, total emails to a daily figure.
meetings per week = 10 positive replies needed = 10 / 0.50 = 20 prospects contacted = 20 / 0.01 = 2,000 per week emails per prospect = 3 (first touch + 2 follow-ups) emails per week = 2,000 x 3 = 6,000 emails per sending day = 6,000 / 5 days = 1,200
Using three emails per prospect is slightly pessimistic, because people who reply or bounce drop out of the sequence. That is fine. Sizing for a little more volume than you strictly need is far cheaper than discovering you are short mid-campaign. If your sequence has more steps, read how to space follow-ups before you multiply, since more steps over a longer window changes the weekly load.
How Much One Mailbox Can Safely Send
This is the number people push hardest, and the one that causes the most damage. A common rule of thumb for cold outreach is roughly 20 to 40 cold emails per mailbox per day once the mailbox is fully warmed. It is not a provider-published limit. It is a level where a single sender still looks like a person doing focused outreach rather than a script.
Plan at the middle of that range, around 30, and treat the top end as a ceiling you reach only after weeks of clean data. Warmup traffic and replies sit on top of cold sends, so a mailbox doing 30 cold emails is doing more than 30 messages a day in total.
How Many Mailboxes per Domain
Domains carry their own reputation, separate from each mailbox. Stacking twenty mailboxes on one domain concentrates all that volume, and all that risk, on a single name. The usual practice is two to four mailboxes per domain, with three a sensible default. At 30 sends each, that keeps any one domain around 90 cold emails a day.
The payoff is containment. If one domain runs into trouble, you lose a small slice of capacity instead of the whole program. That isolation only holds if the layers underneath are separate too, which is why it is worth understanding how IP, domain and mailbox reputation interact.
Three Worked Examples
All three use the same cautious assumptions: 1% positive reply rate, 50% booking rate, three emails per prospect, five sending days, 30 cold sends per mailbox, three mailboxes per domain and about 20% spare capacity.
| Scenario | Meetings per week | New prospects per week | Emails per day | Active mailboxes | With ~20% spare | Domains |
|---|---|---|---|---|---|---|
| Small agency, one client | 3 | 600 | 360 | 12 | 15 | 5 |
| SDR team of four reps | 12 | 2,400 | 1,440 | 48 | 58 | 20 |
| Larger outbound program | 40 | 8,000 | 4,800 | 160 | 192 | 64 |
Two things stand out. First, the infrastructure scales linearly with the goal, so doubling meetings means doubling mailboxes unless your conversion rates improve. Second, the biggest lever is not more mailboxes, it is a better reply rate. Lifting positive replies from 1% to 2% halves every number in the right-hand columns. If you want to run your own inputs, the mailbox calculator does this arithmetic for you.
Build In Warmup Lead Time
New mailboxes on new domains cannot start at 30 cold sends on day one. Plan for at least two to four weeks of warmup before cold volume, followed by a gradual ramp. That means the real deadline for buying domains is roughly a month before you need the meetings, not the week the campaign is approved.
The warmup calculator gives a day-by-day ramp, and the first 30 days of a new sending domain covers what to watch while it runs. Sendbox runs warmup automatically on every connected mailbox, but the calendar time still has to pass. No tool can compress trust that providers build from history.
Spare Capacity for Rotation and Rest
Sizing exactly to the daily target assumes nothing ever breaks. In practice, at any given moment some mailboxes are resting after a dip in placement, some domains are still warming, and occasionally one needs retiring. A buffer of around 15 to 25% above the active count is a reasonable rule of thumb.
Spare capacity also lets you rotate. If a domain's placement drops, you pause it and shift its share of volume to the reserve without pushing healthy mailboxes above their caps. Without a reserve, the only option is to overload the survivors, which is how one bad domain becomes three.
When to Rerun the Numbers
Your first sizing is a forecast. Rerun it whenever an input changes:
- After two to three weeks of sending, replace the placeholder reply and booking rates with real ones.
- When the meeting target changes, for a new client, a new rep or a new quarter.
- When you add or remove sequence steps.
- When placement tests show mailboxes underperforming, since mail landing in spam effectively lowers your reply rate and inflates the volume you think you need.
That last point is the trap. Teams often respond to falling replies by adding volume, when the real problem is that existing volume is not reaching the inbox. Check placement first, then decide whether you need more capacity.
