A good cold email bounce rate in 2026 is under 2%. That's the number mailbox providers use as a rough line between "a sender with normal list hygiene" and "a sender worth watching closely." Cross 5% on any single send and you're not just losing that batch — Gmail and Outlook start quietly routing more of your future mail to spam, regardless of how good the copy is. Bounce rate is the one cold email metric that actively damages every campaign after it, not just the one it's measured on.
Hard bounces vs. soft bounces
The two get lumped together in most reporting, but mailbox providers treat them very differently. A hard bounce means the address doesn't exist — typo, turnover, or a list that was never verified — and it's a permanent, unambiguous signal that you're sending to invalid data. A soft bounce is temporary: a full inbox, a server timeout, a message that was too large. One soft bounce means nothing. A pattern of hard bounces is what actually moves your sender reputation, because it tells Gmail and Outlook you're not checking your list before you send to it.
Why 2% is the real ceiling, not 5%
Ready to turn this into pipeline?
Prospect AI runs research, copy, and multi-channel outreach as one system, so consistent pipeline stops depending on heroics.
5% is the point at which most cold email tools throw a warning, but treating it as the target is backwards — by the time you're there, damage is already compounding. A properly verified list, sent from a warmed domain, should land comfortably under 1%. The 1–2% band is where normal list decay lives: people change jobs, companies get acquired, catch-all domains occasionally reject mail that would otherwise deliver. Sitting under 2% consistently is what keeps a domain in the range mailbox providers consider trustworthy; treating 5% as acceptable is how a domain ends up throttled without anyone noticing until reply rates have already cratered.
What pushes bounce rate up
Almost every high-bounce list traces back to one of three things: no verification step before sending, data that's aged past 3–6 months without a re-check, or scraped contacts pulled from a source that was never designed for outbound in the first place. Catch-all domains are the quiet culprit — they accept mail for any address at the domain regardless of whether that mailbox actually exists, so a verifier can't always catch them, and they show up as bounces only after the send. A list built from a single enrichment pass six months ago is not the same list today; role turnover alone can push 20–30% of a B2B contact list stale within a year.
How bounce rate quietly throttles deliverability
Bounce rate doesn't just cost you the addresses that failed — it's one of the strongest signals mailbox providers use to decide how much of your other mail reaches the inbox at all. A sustained bounce rate above 3–4% is enough for Google Postmaster Tools to flag domain reputation as "bad," and once that happens, sends that would otherwise land in the primary inbox start getting filtered even for prospects with valid, active addresses. This is why teams chasing a weak reply rate often have the wrong problem: the copy might be fine, but a high bounce rate on the list is suppressing inbox placement before anyone reads a word.
Keeping bounce rate under control
The fix is upstream of send, not downstream: verify every address at the point of list-building, not once at signup or import, and re-verify anything older than 60–90 days before it goes into a new sequence. Catch-all domains need a secondary check — pattern-matching against known valid formats, or a lighter-touch first send — rather than being trusted at face value. This is exactly the kind of discipline that erodes under manual process, because verification is the step that gets skipped when a list needs to go out today. An AI SDR that owns the full send pipeline can enforce it every time: verifying before every send, aging out stale contacts automatically, and keeping bounce rate in the low single digits without a rep ever having to remember to run the check.