Soft Bounce vs Hard Bounce: What Each Means and What to Do

Open a cold campaign report and you will see two kinds of failures sitting next to each other: soft bounces and hard bounces. They look similar in the dashboard, but they mean opposite things and demand opposite actions. Treat a hard bounce like a soft one and you keep hammering a dead address until providers flag you as careless. Treat a soft bounce like a hard one and you delete contacts who were only briefly unreachable. This guide explains soft bounce vs hard bounce in plain terms, shows the SMTP codes behind each, and tells you exactly what to do with both so your sender reputation stays intact.

What is the difference between a soft bounce and a hard bounce?

A soft bounce is a temporary delivery failure where the receiving server accepted the connection but could not deliver the message right now, so it may succeed on a retry. A hard bounce is a permanent failure where the server has rejected the message for good, usually because the address does not exist, and no number of retries will ever land it. The simplest way to remember it: soft means "not right now," hard means "never."

The distinction matters because each one calls for a different response and carries a different cost. Soft bounces are normal background noise that your sending tool usually handles for you through automatic retries. Hard bounces are a direct signal to Gmail, Outlook, and Yahoo that your list contains dead data, and every one you fail to remove pushes your email sender reputation down a notch. The line between them is also the line between an address you keep and one you delete.

What causes a soft bounce?

A soft bounce is caused by a temporary, fixable condition on the receiving side: a full mailbox, a server that is down or overloaded, a message that exceeds the size limit, or greylisting, where the server deliberately defers a first-time sender and waits to see if you try again. Rate limiting, where the receiving server throttles how fast it accepts mail from you, is another common cause for cold senders pushing volume.

Because the cause is temporary, the message often gets through on a later attempt. Your mail server queues a soft-bounced message and retries on a schedule, typically after 15 minutes, then 30, then 60, and so on for up to 72 hours before giving up. That retry logic is why you usually do not need to do anything with a single soft bounce. What you do need to watch is a pattern: the same address soft-bouncing across several campaigns, or a sudden spike in soft bounces overall, which often points at your own sending rather than the recipients.

What causes a hard bounce?

A hard bounce is caused by a permanent reason the message can never be delivered: the email address does not exist, the domain is invalid or no longer accepts mail, or the receiving server has blocked you outright. The most common cause in cold outreach is a non-existent address, which comes straight from scraped or stale data where the person has left the company or the address was never real to begin with.

Hard bounces are the dangerous ones for cold senders because they are almost always a data-quality problem you created upstream. An unverified list scraped from the web routinely hard-bounces 15 to 30% of the time, and inbox providers read that as proof you do not check your data before sending. Retrying a hard bounce does nothing except generate more rejections, so the only correct response is to remove the address permanently and fix where the bad data came from. That is why list hygiene matters: clean data is the difference between a 1% bounce rate and a domain-burning 20% one.

What are soft bounce and hard bounce SMTP codes?

Bounces map to SMTP response codes, and the first digit tells you which kind you have: 4xx codes are temporary failures (soft bounces) and 5xx codes are permanent failures (hard bounces). A code like 421 or 450 means try again later; a code like 550 or 553 means the message was permanently rejected. Reading the code is the fastest way to know whether to retry an address or remove it.

One nuance worth knowing: a repeated 4xx that never resolves effectively becomes a hard bounce. If your server retries a soft-bounced address for the full 72-hour window and still cannot deliver, most tools convert it to a permanent failure and stop. So a single soft bounce is harmless, but the same soft bounce campaign after campaign should be treated like a hard one and pulled from your list.

What should you do with a soft bounce?

For a single soft bounce, do nothing: your sending tool queues the message and retries it automatically over the next few hours, and it usually lands. The temporary condition that caused it (a full inbox, a busy server, greylisting on a first contact) tends to clear on its own, so manual intervention is rarely needed and deleting the address would lose you a valid contact.

What you should act on is the pattern, not the single event. If an address soft-bounces across three or four consecutive campaigns, treat it as unreachable and remove it. If your overall soft-bounce rate jumps suddenly, the cause is usually on your side: you are sending too fast and getting throttled, or your sending reputation has dipped and servers are deferring you. The fix there is to slow down, check that your domain is warmed, and confirm your authentication is intact rather than blaming the recipients.

What should you do with a hard bounce?

Remove a hard-bounced address from your list immediately and permanently, ideally with automatic suppression so it can never be sent to again. There is no recovery path: the address does not exist or has rejected you for good, and every additional send to it produces another rejection that providers count against you. The faster you suppress hard bounces, the less damage they do.

Then fix the source. A hard bounce is a symptom of bad data, so a high hard-bounce rate means your list was scraped, stale, or never verified. The cure is upstream: verify every address before you send and re-check any list older than 90 days. Reading how to clean an email list alongside how to build a cold email list covers the exact process that keeps hard bounces near zero in the first place, which is far cheaper than cleaning up after them.

What is a good bounce rate for cold email?

A good total bounce rate for cold email is under 2%, the same threshold inbox providers use, and crossing 5% means you should stop sending and clean your list before the next campaign. That 2% figure covers both bounce types combined, though for cold outreach the number that really tells the story is your hard-bounce rate, since a high one is a direct verdict on your data quality.

Staying under 2% is almost entirely a list-quality outcome. A verified list bounces under 2% because the addresses are real; an unverified scraped list bounces 15 to 30% because much of it never existed. If your bounce rate is climbing, the fix is in your data and your sending discipline, not your copy. The same principles run through everything in cold email bounce rate, which goes deeper on the targets and how to read them.

How do you prevent bounces in the first place?

You prevent bounces by verifying your list before every send and warming your domain before you scale. Verification removes the invalid addresses that cause hard bounces, and a warmed, authenticated domain stops the throttling and deferrals that cause soft bounces. Do both and your combined bounce rate stays comfortably under the 2% line without much ongoing effort.

The supporting habits matter too: send at a steady pace rather than in spikes so receiving servers do not defer you, keep SPF, DKIM, and DMARC in place so providers trust your mail, and suppress every bounce after each campaign so your list only gets cleaner over time. These are the same fundamentals that drive the rest of cold email deliverability, and they are why disciplined senders see near-zero bounces while careless ones burn domains. For the warmup side specifically, how to warm up an email domain walks through the ramp that keeps soft bounces down on a new domain.

The short version

A soft bounce is a temporary failure (a 4xx SMTP code) that your sending tool retries automatically, so a single one needs no action; only remove an address if it soft-bounces repeatedly across campaigns. A hard bounce is a permanent failure (a 5xx code, usually a non-existent address) that you must suppress immediately, because every send to it damages your reputation. Keep your total bounce rate under 2% and stop sending above 5%. The way to hit those numbers is upstream: verify your list before you send and warm your domain before you scale, which prevents both bounce types far more cheaply than cleaning up the fallout.

Once your bounces are under control and replies start landing, you will want those responses and bounce notifications sorted automatically rather than by hand; an email parsing tool turns reply and bounce emails into structured CRM data for you. If email is one channel in a wider mix, a WhatsApp bulk messaging platform gives you a second route to the same verified prospects. And if you would rather your inbound content engine run on its own while you focus on outbound, an AI SEO agent handles the blog side on autopilot.