A soft bounce is a temporary delivery failure: the address is real, but the message could not land right now. A hard bounce is permanent: the address, the domain or the recipient’s server will never accept the message. You retry the first and suppress the second. For cold email, where every send is judged on sender reputation, getting that call right is part of keeping your domains alive. PlusVibe is built around inbox placement, and bounce hygiene is the first layer of it.
Key Takeaways
- A soft bounce returns a 4xx code and means “try again later”; a hard bounce returns a 5xx code and means “stop sending to this address.”
- Suppress a hard-bounced address after one failure. Retrying it tells mailbox providers you do not clean your list.
- Treat a soft bounce as a warning only when it repeats: the same address failing on about three sends over several days should be pulled from the sequence.
- Catch-all domains accept everything, so they hide bad addresses until after you send. Verify, segment and send them slowly.
- Keep hard bounces under 2% of sends. Cold lists above that need to be paused and re-verified.
Table of Contents
- What a bounce actually is
- Soft bounce vs hard bounce at a glance
- What causes a soft bounce
- What causes a hard bounce
- How to read a bounce message
- What to do after a soft bounce
- What to do after a hard bounce
- Why retrying a hard bounce hurts you
- What bounce rate is acceptable
- Catch-all emails: the bounce you cannot see coming
- A bounce routine for outbound teams
- Our view: bounces are the first thing that tells you the truth
- Keep bounces low with PlusVibe
- FAQ
- Sources
What a bounce actually is
A bounce is a message sent back by a receiving mail server (or your own) saying it did not deliver your email. The message includes a status code from SMTP, the protocol servers use to hand mail to each other. The first digit of that code carries the verdict: 4 means a temporary condition, 5 means a permanent one.
That single digit is the whole difference between the two bounce types. Everything else, the cause, the fix and the damage to your reputation, follows from it.
Soft bounce vs hard bounce at a glance
| Soft bounce | Hard bounce | |
|---|---|---|
| Meaning | Temporary failure | Permanent failure |
| SMTP code family | 4xx | 5xx |
| Typical causes | Full mailbox, server down, rate limiting, greylisting, reputation deferral | Address does not exist, domain does not exist, recipient server blocks you |
| Does the address exist? | Usually yes | Usually no |
| What to do | Retry automatically, then pull it if it keeps failing | Suppress immediately |
| Risk if ignored | Slow reputation drag | Fast reputation damage, spamtrap exposure |
What causes a soft bounce
Soft bounces come from conditions on the receiving side that may clear up. Spamhaus, the nonprofit that runs widely used blocklists, lists the common temporary codes:
- 421: the service is unavailable or connection limits were hit.
- 450: the mailbox is unavailable, for example corrupted, offline, or deferred because of reputation or a blocklist.
- 451: a local processing error on the receiving server.
- 452: too many emails were sent, or a storage limit was exceeded.
In plain terms: the inbox is full, the server is busy or down, the recipient’s system is rate limiting you, or it is greylisting you (deliberately rejecting a first-time sender once and expecting a retry). Your sending system should keep retrying until the message is accepted or its own timeout runs out.
Two of these deserve attention in outbound. A 450 or 421 that mentions reputation or rate limits is not about the recipient at all. It is the receiving server saying it does not trust your sending right now. If soft bounces cluster at one provider, such as Gmail or Outlook, look at your sending volume and your domain health before you blame the list.
What causes a hard bounce
Hard bounces are final. Spamhaus describes them as permanent rejections with 5xx codes, and lists examples: 550 for a non-existent address, 512 when the domain cannot be found in DNS, 551 for relay denied, 552 for storage exceeded and 553 for an invalid mailbox name.
In practice the causes are:
- The person left the company and the mailbox was deleted.
- The address has a typo, or your enrichment tool guessed the format wrong.
- The domain does not exist or has expired.
- The recipient’s server blocks your sender outright.
The first three are list-quality problems. The last one is a reputation problem, and it needs a different fix: look at your authentication records and sending patterns, not the address.
Spamhaus is blunt about the rule: never send again to a hard-bounced contact. It adds a reason that matters for cold outreach. Non-existent domains can be turned into spamtraps, addresses that exist only to catch senders who mail unverified lists. Hitting one is a strong signal to blocklist operators that your list was bought, scraped or neglected.
How to read a bounce message
The code alone is not always enough, so read three things in the bounce notice.
- The status code. A 4.x.x enhanced code is temporary, a 5.x.x code is permanent. Common examples are 5.1.1 (mailbox does not exist) and 4.2.2 (mailbox full).
- The human-readable text. It often says whether the problem is the recipient (“user unknown”) or you (“blocked due to reputation,” “policy violation”).
- Who sent the bounce. If it came from the recipient’s server, the problem is on their side or about your standing with them. If it came from your own outbound system, it may be a DNS, authentication or configuration problem.
Bounce types are not always clean. Nylas documents an edge case where a permanently over-quota mailbox is graded as a hard bounce with a 5.2.2 code, even though “full mailbox” sounds temporary. Trust the code over the description, and treat a persistent “full” as dead.
For cold email, a spike matters more than one message. If your bounce rate jumps overnight, our guide to bounce spikes and the seven fastest root-cause checks walks through where to look first.
What to do after a soft bounce
Do nothing at first. Soft bounces are normal, and a single one means very little. Most sending systems retry on their own with increasing gaps between attempts.
The rule that matters is what to do when the same address keeps failing. Nylas recommends keeping a soft-bounced address active, but suppressing it once it soft-bounces on three consecutive sends spread across at least 72 hours. The spacing is the point: three failures in one afternoon might be one outage, three across three days is a pattern.
Suppressing after one soft bounce is the opposite mistake. You discard a real prospect whose mailbox was briefly full. Suppressing too late is the other. Three strikes across at least 72 hours is a reasonable middle.
When soft bounces hit many addresses at once rather than one, stop and diagnose:
- Are they all at one mailbox provider? Slow down volume to that provider.
- Do the messages mention reputation or blocklists? Check your domain and IP status.
- Did you just add a new domain or mailbox? A new sender that ramps too fast gets deferred. A proper warm-up avoids this.
What to do after a hard bounce
Suppress the address after the first hard bounce. A suppression list is simply a record of addresses your system must never email again, and it should apply everywhere you send, not only in one campaign.
Two habits cover most of the risk:
- Apply suppression across every tool. If an address hard-bounces in a sequence and your CRM still lists it as active, the next campaign will mail it again.
- Do not re-add an address unless you have confirmed it is valid. An address that was invalid last quarter is rarely fixed by waiting.
The word “delete” tempts people, but suppression is the safer habit. A deleted address can be re-imported next month from a fresh scrape and mailed again. A suppressed one stays blocked.
How other systems handle this shows the logic. Cloudflare’s email documentation says hard bounces are never retried, while soft bounces are retried with exponential backoff, meaning longer and longer waits between attempts. Its suppression list rules keep a hard-bounce suppression permanently when the mailbox or domain does not exist, and spam-complaint suppressions never expire. The same principle applies to your outreach stack: permanent failures stay blocked.
Why retrying a hard bounce hurts you
Mailbox providers watch how senders behave after a failure. Repeated attempts to an address that has already returned a permanent error look like list-hygiene neglect, which is also what spammers look like. The cost is not one lost email. It is a lower sender reputation that pushes later, perfectly good messages toward spam folders.
This is also why list quality starts before the first send. Building a list from people who actually want to hear from you keeps invalid addresses out from the start, as these ways to grow an email list show. For outbound, where lists are built rather than opted in, verification does that job.
What bounce rate is acceptable
There is no universal number, but there are useful ceilings. Cloudflare’s deliverability guidance sets a delivery rate above 95%, a hard bounce rate below 2% and a complaint rate below 0.1% as the marks of a healthy sender.
For cold email, treat 2% hard bounces as the line, not the target. Our write-up on getting cold email bounces under 2% goes through the changes that move the number, and the acceptable bounce rate thresholds page covers the benchmarks in more depth.
A worked example. You send 5,000 emails in a week. 60 hard-bounce: that is 1.2%, within range, but suppress all 60. Next week, 5,000 more, and 190 hard-bounce: that is 3.8%. Pause the campaign, find out which list segment those 190 came from, and re-verify it before you send another message. A 20% bounce rate, which readers sometimes ask about, means the list itself is the problem. Stop sending and rebuild it.
Track hard and soft bounces as separate numbers. A blended figure can hide a clean list with a temporary outage, or a healthy-looking total with a dying list underneath.
Catch-all emails: the bounce you cannot see coming
A catch-all domain is configured so its mail server accepts messages sent to any address at that domain, whether or not a real mailbox exists. For example, anything@company.com is accepted, even if nobody is called Anything.
This is a problem for verification. Normal checks ask the server whether an address exists. A catch-all server says yes to everything, so a verifier cannot confirm a specific mailbox and usually labels it “unknown” or “accept-all” rather than valid or invalid. Catch-all is a property of how the domain is set up, so it is not a flaw of any one verifier.
The risk shows up after sending. The server may accept your message and then bounce it later, or route it to a mailbox nobody reads. That is why catch-all addresses are a common source of delayed bounces.
How to detect a catch-all domain
The standard test is to probe the domain with an obviously fake address. If the server accepts it, the domain is flagged as catch-all. That tells you about the domain only. It says nothing about whether your contact’s mailbox is real.
How to handle catch-all contacts in cold outreach
You have three options, in order of safety:
- Leave them out when your list is large enough that you can afford to. This is the cleanest route for a new domain or a freshly warmed mailbox.
- Segment them. Keep catch-all contacts in their own list, so their bounces do not blur into the rest of your numbers.
- Send slowly. Use smaller batches, lower daily volume and stricter monitoring, and stop the segment if bounces or complaints appear.
Then follow the results. Suppress any catch-all address that bounces, and consider dropping those that never open or reply after several touches.
Because hard bounces from catch-all domains can arrive late, check the segment again a day or two after each batch rather than only at the time of sending. If you want to verify addresses before they enter a sequence, our email verification covers that step, and the free email verification tool lets you check addresses one at a time.
A bounce routine for outbound teams
A short routine keeps the numbers healthy without turning bounce handling into a project.
- Verify before you load. Run every new list through verification, and separate valid, catch-all and invalid results.
- Warm new mailboxes. New senders that send full volume on day one get soft-bounced and deferred. Our email warm-up ramps volume gradually in a private pool.
- Send in small batches first. Watch the first few hundred messages of a new list before releasing the rest.
- Suppress hard bounces on the first failure and soft bounces after the same address fails about three times over three days or more.
- Review weekly. A 30-minute weekly deliverability audit is enough to catch a drifting bounce rate before it becomes a reputation problem.
- Clean the list regularly. Our guides on list hygiene and email list cleaning cover the schedule.
If you want the full picture of what a bounce does after it happens, what to do when an email bounces back goes through recovery step by step.
Our view: bounces are the first thing that tells you the truth
We think of bounce data as the earliest honest feedback in outbound. Open rates are fuzzy and reply rates arrive late, but a 550 tells you immediately that a list was wrong.
That is why we build PlusVibe around the inbox rather than the send button. Each agency client can have its own sending server and dedicated IPs on the Agency plan, so one client’s bad list never touches another client’s reputation. Mailboxes warm up in one closed private pool, with bad senders removed and every inbox re-checked over time, and its peak warm-up bounce rate has never exceeded 0.1%. Seed tests then show where your real campaign emails land at Gmail, Outlook and other providers, so you are not guessing from a dashboard score.
Our rule is simple: outreach should be relevant, honest, respectful and measured. Suppressing a hard bounce on the first failure is the measured part. Receipts over vibes.
Yaro
Keep bounces low with PlusVibe
If you send enough email that a bounce rate of a few percentage points means hundreds of lost contacts and damaged domains, the infrastructure matters as much as the list. PlusVibe gives you verification, a private warm-up pool, sequencing and seed-test placement checks in one platform. Plans start at $37/month, Business is $77/month for 150,000 emails and Agency starts at $497/month for 500,000+ emails, with no fee per connected mailbox.
You can start the 14-day free trial with no credit card and judge it on placement. See the plans or read how our deliverability features work.
FAQ
Should I delete hard bounce emails?
Suppress them rather than only deleting them. A suppression list keeps the address blocked in every campaign and tool, so a later import or scrape cannot bring it back. Never send to a hard-bounced address again, because repeated attempts damage your sender reputation.
How many soft bounces before I remove an address?
One soft bounce means almost nothing. A common rule, described by Nylas, is to suppress an address after three consecutive soft bounces spread over at least 72 hours. If many addresses at one provider soft-bounce together, check your sending volume and reputation instead.
Can a soft bounce turn into a hard bounce?
In effect, yes. If an address keeps failing with temporary errors, most systems stop retrying and treat it as undeliverable. A mailbox that is full for weeks can also be graded as a permanent failure by the receiving server, using a 5.2.2 code.
Is it safe to email catch-all addresses?
It carries more risk than a verified address, because the server accepts the message whether or not the mailbox exists. If you include catch-all contacts, put them in a separate segment, send in small batches, and suppress any that bounce.
Sources
- Spamhaus: How to handle bounced emails
- Nylas: Soft bounce vs hard bounce explained
- Cloudflare Email Service: Email deliverability
- Cloudflare Email Service: Suppression lists
- How to grow an email list: proven strategies


.jpeg)

























































