How many cold emails per day is actually safe

"The more I send, the more replies I get" is intuitive, and in cold email it is close to the opposite of the truth.

The per-mailbox ceiling

A warmed mailbox can sustain a few dozen cold emails a day. A new one should start at roughly 10 to 20 while it is being warmed up. The limit is not the tool, it is what mailbox providers consider normal human behaviour.

Scale sideways, not up

The way to increase reach is more warmed mailboxes each sending a safe amount, not one mailbox sending more. Total volume grows and every individual sender stays under the threshold.

Each new mailbox needs its own warm-up before joining a campaign. Skipping that reintroduces exactly the risk you were avoiding.

Why more is usually worse

High volume without tight targeting means more mediocre messages going to people who do not fit. That drives down reply rates and drives up bounces and complaints, which mailbox providers read as a signal about you.

So the damage is double: worse campaign results now, worse deliverability later.

A safe sending ramp, week by week

Sending volume is earned, not declared. A new domain and mailbox pair has no history, and mailbox providers treat unknown senders with suspicion until behaviour proves otherwise. A ramp that respects that looks like this:

1. Weeks 1 to 2: warm-up only. No campaigns. The mailbox exchanges low-volume, natural-looking mail so providers see engagement before they see outreach.
2. Weeks 3 to 4: first campaign sends. Start around 10 to 20 cold emails a day per mailbox, on top of continued warm-up activity.
3. Weeks 5 to 6: gradual increase. Add a handful of sends per day each week, watching bounces and replies as you go.
4. Week 7 onwards: cruising level. Settle at a few dozen a day per mailbox and hold there. This is the ceiling, not a checkpoint.

If deliverability wobbles at any stage, drop the volume back a step and hold until it recovers. Pushing through a wobble is how domains get burned.

Working out how many mailboxes you need

Once the per-mailbox ceiling is fixed, capacity becomes arithmetic. Say you want to reach 1,000 prospects a month with a four-email sequence. That is roughly 4,000 sends, or about 200 per working day. At a safe few dozen sends per mailbox, that means somewhere between six and ten warmed mailboxes, spread across two or three separate sending domains so no single domain carries the whole campaign.

Most teams get this backwards: they pick a volume target first and hope the infrastructure copes. Plan it the other way, from mailboxes up. The full method, including how to phase new domains in, is in how many mailboxes you need.

Signals you are already sending too much

Sender reputation degrades quietly before it collapses. Watch for these:

• A bounce rate that keeps climbing week on week. Providers read hard bounces as evidence of a scraped or guessed list, so the trend matters more than any single figure.
• Spam complaints approaching the 0.3% threshold. Past it, bulk sender rules make filtering close to automatic.
• Reply rate falling while volume holds steady. The most common early sign that more mail is landing in spam.
• Warm-up tools reporting more messages rescued from junk folders. A direct read on where your mail is landing.

Any one of these is a reason to pause and investigate, not to push harder. Ongoing checks are covered in monitoring deliverability.

Common sending volume mistakes

• Blasting a fresh domain. Hundreds of sends in week one is the fastest way to lose a domain. Fix: warm up first, always.
• Treating the tool's limit as the safe limit. Sending platforms let you schedule far more than providers tolerate. Fix: set your own caps below the tool's.
• Ramping every mailbox at once. If something goes wrong you cannot tell which variable caused it. Fix: stagger new mailboxes by a week or two.
• Ignoring send spacing. Fifty emails fired in one minute looks like a machine. Fix: spread sends across the working day with randomised gaps.
• Counting follow-ups as free. Every step in a sequence is a send against the daily cap. Fix: budget the full sequence, not just email one.

"But I see agencies sending thousands a day"

They are, across dozens or hundreds of mailboxes, each one held under the same per-mailbox ceiling described here. The headline number is an aggregate, not a trick that bypasses provider limits. Some volume shops do burn domains and treat them as disposable; that model produces terrible reply rates and drags the sender's brand through spam folders. It is not a model worth copying, and it does not survive contact with a niche B2B market where the list of relevant companies is finite.

How this plays out in European markets

In the US, a broad ICP can justify big infrastructure because the addressable list runs to tens of thousands of companies. In most European niches it does not. A campaign into the DACH region or the Baltics often targets a few hundred to a few thousand relevant firms in total, which means the constraint on volume is the market, not the mailbox count. Burning through that finite list with high-volume, low-relevance sends is unrecoverable in a way it never is in a huge market. GDPR adds a second reason for restraint: legitimate interest is far easier to defend when outreach is targeted and modest than when it is indiscriminate, and German law sets a higher bar again.

Let list quality set the volume

The better question is not "what is my maximum?" but "how many genuinely relevant contacts do I have?" Volume should follow list quality. Reaching 200 well-matched companies properly beats spraying 2,000 badly.