When to Pause a Cold Email Campaign Before Bounces Hurt Your Domain

October 29, 2026 · 3 min read

Warming up a new sending domain gets most of the attention, because it's a thing you do once, on a schedule, with a clear end point. Watching bounce rate once a campaign is actually running gets less, because it's ongoing and nothing marks the moment it starts to matter. That's backwards: a warmed-up domain can still get hurt badly by a bad list two months in.

Why a hard bounce is worse than it looks

A hard bounce (the address doesn't exist, the domain is dead) tells a receiving mail server something specific: whoever sent this didn't check their list. One or two in a batch of hundreds is normal — phone numbers go stale, so do business email addresses. A rate climbing over a small handful of percent in a short run is a different signal entirely, and mail providers read it as evidence worth acting on, not noise.

The part that catches people out: the damage isn't to the one bounced message. It's to how your sending domain gets treated afterward, for every message sent from it for a while — including the ones that would otherwise have landed fine.

A rule of thumb that works without a dashboard

You don't need sophisticated monitoring to catch this early, just a minimum sample and a threshold:

Don't read anything into a bounce rate from a tiny batch. Three bounces out of five sent is meaningless; the same three out of five hundred is not. Wait for at least twenty or so sent before drawing a conclusion.
A rate creeping past roughly 3% is worth a look — not an emergency, but worth checking where the list came from before it climbs further.
A rate at or past roughly 5% is past the point of "worth a look." At that point, the safer move is to stop sending from that domain until the list is cleaned, not to keep going and hope it was a bad batch.

Those aren't universal numbers mail providers publish anywhere — they're a reasonable, conservative line to draw given how quickly reputation damage compounds, and they're close to what AyePin itself uses as defaults, described below.

What actually causes the spike, most of the time

Almost never the sending setup itself, if that was warmed up properly. Almost always the list: a batch of addresses scraped from an old directory, guessed rather than confirmed, or imported from a source that was stale six months before you got it. The fix is upstream of the sending tool — checking addresses before a campaign starts, not after bounces are already climbing.

How AyePin handles this automatically

AyePin watches a rolling window of the last fifty sent emails per workspace, and won't draw a conclusion from fewer than twenty of them. Past a 3% hard-bounce rate in that window it starts surfacing a warning; past 5%, it pauses sending on that channel automatically rather than letting a bad list keep running unattended. Both thresholds can be tuned per install, but the defaults are deliberately conservative.