Target under two percent hard bounces. Two to five percent is where providers start treating the list as bought or stale. Above five percent, pause the campaign: every further send is adding evidence against a domain you spent weeks warming.
Published by emailcampaign.ai, which runs sending infrastructure and warms mailboxes. The thresholds below are the ones we act on in our own sending.
The thresholds
| Hard bounce rate | How receivers read it | What to do |
|---|---|---|
| Under 2% | Normal list decay | Continue |
| 2% to 5% | List looks bought or stale | Slow down, re-verify before the next batch |
| 5% to 10% | Strong negative signal against the domain | Stop. Re-verify everything before resuming |
| Over 10% | The list is not usable | Stop and rebuild the list. Expect weeks of recovery |
These are not arbitrary. Hard bounce rate is the cheapest proxy a receiver has for whether you obtained addresses from people or from a file, and it is evaluated against the sending domain, which is why one bad campaign affects the next good one.
Hard and soft are different measurements
A hard bounce is permanent: the address does not exist. That is a statement about your list. A soft bounce is temporary: a full mailbox, a throttle, a server having a bad afternoon. That is a statement about conditions.
Blending them produces a number that means nothing, and it fails in both directions. It inflates the figure you are judged on, and it hides genuine list problems behind transient noise. Check whether your tool separates them. If it reports a single bounce rate, find out which codes it counts before trusting any threshold, including the ones on this page.
The classification that matters is the enhanced status code rather than the three digit code, and two cases are routinely misread: mailbox full (5.2.2) is formatted as permanent but behaves as temporary, and policy rejections (5.7.x) are about the sender rather than the recipient. Suppressing on either loses you reachable people.
Why the list is nearly always the cause
B2B data decays quickly as people change jobs, and a list assembled a year ago will bounce even if it was accurate when built. The high-bounce campaigns we see fall into four groups:
- Purchased or scraped files, which were aggregated rather than verified and carry spam traps as well as dead addresses.
- Guessed address patterns, built by combining a name with a domain. The pattern is right often enough to feel like it works and wrong often enough to bounce.
- Lists that sat unused, where nothing was wrong at the time of collection.
- Bulk sends to catch-all domains, where a verifier reported valid because the server accepts everything, and the failure arrives later as an asynchronous bounce.
Keeping it under two percent
Verify close to the send. Verification ages. A file checked three months ago is not a verified file today.
Bucket accept-all separately. On a catch-all domain, valid means only that the server did not say no. Send to that bucket last, in smaller batches, and watch what comes back.
Batch and check. Send a few hundred, read the bounces, then continue. Discovering a bad segment after twenty thousand sends is a different outcome from discovering it after three hundred.
Suppress permanently and across campaigns. A hard bounce that reappears in next quarter’s list is an avoidable repeat.
If it is already running hot
Pause. The instinct to finish the campaign is the expensive one, because the reputation cost accrues per send and the recovery is measured in weeks of clean sending rather than days.
Then: re-verify what remains, split out the accept-all addresses, resume at a lower per-mailbox volume, and check between batches. If the domain was new, expect to spend time rebuilding history before returning to normal volume.
Questions, answered straight
- What is an acceptable bounce rate for cold email?
- Under two percent hard bounces is the working target. Between two and five percent, providers begin reading the list as bought or stale. Above five percent you should stop the campaign rather than finish it, because the damage attaches to the sending domain rather than to the campaign.
- Do soft bounces count toward that number?
- Not in the same way. Hard bounces indicate list quality, which is what receivers judge. Soft bounces indicate temporary conditions such as a full mailbox or a throttle. Tools that report one blended bounce rate make the number useless, which is a good reason to check how yours is calculated.
- What causes a high bounce rate?
- Usually list sourcing. Purchased and scraped lists decay and were often assembled without verification. Other causes are sending to catch-all domains in bulk, a list that has sat unused for a year, and address patterns guessed from a name and a domain rather than verified.
- How do I lower a bounce rate mid-campaign?
- Pause rather than push through. Re-verify the remaining addresses, separate accept-all results into their own segment, resume at lower volume per mailbox, and check bounces between batches instead of after the campaign.
- Does a high bounce rate damage the domain permanently?
- Not permanently, but recovery is slow and is measured in weeks of good behaviour rather than days. That asymmetry is why stopping early costs far less than finishing a campaign that is bouncing.
Published 20 September 2026. Updated 20 September 2026. Written by the team that runs the infrastructure; numbers come from the platform's own provisioning and sending, and from the providers' published documentation at the time of writing.