Soft bounce

How a soft bounce works

Soft bounces sit on the temporary side of the bounce taxonomy. AWS lists examples such as a full mailbox, too many connections, also called throttling, or a connection that times out 2. The key property is that the failure may clear without changing the address. A mailbox that is full on Monday can accept mail on Wednesday.

Amazon SES retries soft bounces multiple times 2. Other ESPs do the same with their own windows and caps. If retries keep failing, some systems eventually convert the event into a permanent failure. That conversion is an operational choice, not proof that the mailbox never existed. Operators should read the bounce code and the retry policy before they suppress.

AWS also notes that soft bounces are not permanent failures or rejections in the same sense as hard bounces 1. Treating every soft bounce like a hard bounce deletes contacts who would have accepted the next attempt. Treating every hard bounce like a soft bounce keeps dead addresses in the retry queue and inflates bounce rate.

Soft bounces are also where sending pace shows up. Hitting a provider with too many connections triggers throttling 2. The address may be fine. The sending pattern is not. Warm-up schedules, per-domain rate limits, and separate IPs for risky traffic reduce that class of soft bounce.

Verification cannot eliminate soft bounces. A mailbox can verify clean and still be full at send time. What email verification removes are the permanent failures that would have looked like noise in the same report 3. RichAPI Email verifier returns mxFound, disposable, valid, and status so those permanent cases leave the list first 3.

Reading soft-bounce codes pays off. Mailbox full is a recipient-side capacity issue. Throttling is often a sender-side pace issue 2. Timeouts can be either network noise or a receiver under load. Each code suggests a different tweak. One shared suppress list for all of them destroys that signal.

Some ESPs eventually promote a repeated soft bounce into a permanent failure. That promotion is useful when it is documented. It is harmful when it is silent and a recoverable contact disappears from every audience. Ask the ESP how many soft failures trigger promotion, and mirror that rule in the CRM so sales tools do not re-import the address from another system.

Soft bounces also cluster during warm-up. A new sending domain that jumps to full volume will see temporary rejects even on clean addresses. That pattern is a pacing problem. Verification still helps by removing permanent failures from the same window so the warm-up chart stays readable 3.

Why soft bounces matter for outbound

Soft bounces are easy to misread. A dashboard that shows only bounce rate mixes temporary throttling with dead mailboxes. Teams then either over-suppress or under-suppress. Split the labels. Retry soft bounces on a schedule. Suppress hard bounces the same day.

Volume plans should expect some soft bounces even on clean lists. Mailboxes fill up. Corporate filters throttle unknown senders. Shared ESP IPs absorb neighbor traffic. The response is pacing and authentication, not a panic purge of every temporary failure.

Where RichAPI fits is upstream. Soft-bounce noise is harder to interpret when the list still contains domains with no MX and disposable inboxes. Cleaning those with email_verifier before the campaign makes the remaining soft bounces more actionable 3. Each verdict costs 2 credits 3. A miss with no verdict costs zero 3. New accounts get 100 free credits 3.

Agencies should also separate soft bounces by client and by sending domain. One client's aggressive cadence can throttle a shared IP and create soft bounces on another client's mail. Attribution prevents the wrong team from rewriting copy for a pacing problem.

After several soft bounces to the same address across days, escalate. A mailbox that never accepts mail starts to behave like a permanent failure even if the first code looked temporary. Document that threshold so suppression stays consistent across the team.

Content teams sometimes chase soft bounces with subject-line tests. That can help when the real issue is spam complaints. It does nothing for a full mailbox. Separate the metrics. Soft-bounce rate, complaint rate, and hard-bounce rate answer different questions for email deliverability and need different owners.

For agencies, soft-bounce spikes after a shared-IP neighbor sends a bad campaign are common. The fix may be IP separation, not a rewrite of every client's first line. Deliverability reviews should include neighbor traffic and authentication status beside list quality.

RichAPI's role stays upstream of that review. email_verifier removes domains that cannot receive mail and labels disposable domains before the ESP ever sees them 3. The call costs 2 credits per verdict 3. A no-verdict miss costs zero 3. New accounts get 100 free credits 3. Soft-bounce retries remain an ESP concern because RichAPI does not send mail.

Example of handling a soft bounce

Suppose a first wave sends to five hundred contacts and thirty return soft bounces with mailbox-full or throttling codes. The ESP retries those thirty over the next day 2. Twenty deliver on retry. Ten still fail.

The operator leaves the twenty in the active audience. The ten move to a watch list. If they soft bounce again on the next campaign, they join the suppression table. They do not join it on the first temporary code alone.

Before a larger second wave, the team verifies new imports with email_verifier 3. Permanent failures leave early through mxFound false, disposable true, or invalid status 3. That way the soft-bounce report on wave two is mostly pacing and mailbox state, not dead domains that would have inflated bounce rate.

Credits on the verify pass stay predictable. Verdicts bill 2 credits 3. No-verdict misses bill zero 3. A new account can run a first pass against the free 100 credits 3. The soft-bounce handling itself stays in the ESP. RichAPI does not send mail and does not retry bounces.

Another example is a warm-up schedule. A new domain starts at fifty sends a day and climbs slowly. Soft bounces during that climb trigger a pause at the current rung rather than a jump back to zero. Hard bounces during the same climb trigger suppression and a list check, rather than only a slower ramp.

Before each rung, new addresses pass email_verifier 3. Permanent failures never enter the warm-up pool 3. That keeps the soft-bounce chart about pace and recipient capacity. Credits stay at 2 per verdict, zero on a no-verdict miss, with 100 free credits available on a new account 3.

When soft bounces concentrate on one corporate domain, the operator checks whether that domain throttles unknown senders. Authentication, a slower cadence to that domain, and a smaller daily cap are the levers. Deleting every soft-bounced contact at that company would erase people who would accept mail a week later.

Soft bounce vs hard bounce

Soft and hard bounces are both delivery failures. Soft means temporary. Hard means persistent. AWS documents the split and the different retry behavior 12.