The AWS SES suppression list is not enough to protect your sender reputation
Learn what the Amazon SES suppression list covers, how bounce events differ from reputation metrics, and how to investigate repeated delivery failures.
The Amazon SES account-level suppression list is an important safeguard, but it does not replace your application’s consent records, delivery diagnostics, or retry policy. A recipient can have repeated delivery problems without appearing on that list. The right response depends on the failure: a nonexistent mailbox, a full inbox, and a rejected attachment need different fixes.
This guide explains what SES suppression does, how to distinguish raw bounce events from reputation metrics, and where an application-level review process helps.
Correction, October 14, 2026: An earlier version misstated suppression retention, bounce classifications, and AWS reputation metrics. It also treated Reject events as proof of invalid recipients. Those claims have been corrected; the account example and conclusions that depended on the incorrect metric interpretation have been removed.
What the SES suppression list does
The account-level list belongs to an AWS account and Region. Check the settings in every Region you send from, including any configuration-set overrides. Automatic additions depend on your enabled suppression reasons: BOUNCE, COMPLAINT, or both. SES automatically adds hard-bounce recipients when bounce suppression is enabled; this is not limited to the Permanent/General subtype.
An entry remains until you remove it. There is no default 14-day expiry or configurable 365-day retention period. AWS documents an exception: after an account’s sending ability has been paused for 90 days, its account-level suppression entries are deleted.
AWS provides console and API tools to view, add, and remove addresses. See the account-level suppression documentation for configuration and removal procedures.
Raw bounce events are not the reputation bounce rate
Do not label every ratio of bounce events to sends as AWS’s reputation metric.
When account-level suppression prevents delivery, the attempt is excluded from Reputation.BounceRate and Reputation.ComplaintRate. It can still appear in the raw Bounce or Complaint metrics and consume daily sending quota. That is a reason to check suppression before sending, but it is not evidence that AWS penalizes every suppressed attempt in its reputation rate.
Use the SES reputation dashboard and its documented metrics for account-health monitoring. Use your event stream for a different question: which messages failed, why, and what should change before the next attempt? AWS explains the distinction in its reputation monitoring guide.
For an internal report, record the numerator, denominator, time window, and whether you count recipients, messages, or events. Deduplicate repeated event deliveries. Keep permanent, transient, and suppression-related failures separate so an operational failure ratio cannot be mistaken for an AWS enforcement metric.
Read the failure before deciding what to suppress
SES’s bounce notification reference describes the types and subtypes. Use those fields together with the recipient diagnostic rather than inferring a cause from the subtype alone.
| Signal | What to investigate | Suggested response |
|---|---|---|
Permanent/NoEmail or Permanent/General | Invalid address or another permanent delivery failure | Stop routine retries; verify the cause before restoring delivery. |
Permanent/OnAccountSuppressionList | Account-level suppression prevented delivery | Inspect the existing suppression reason and fix the application’s repeated attempts. |
Transient/MailboxFull | Recipient mailbox capacity | Apply a bounded retry policy; review repeated failures. |
Transient/MessageTooLarge, ContentRejected, or AttachmentRejected | Message size or content constraints | Fix the message before deciding the recipient is invalid. |
Transient/General or Undetermined/Undetermined | A failure without a sufficiently specific classification | Inspect diagnostics, recurrence, and any later successful deliveries. |
Reject | SES refused the message before delivery | Inspect the rejection reason and message content; this does not establish that the mailbox is dead. |
A Reject is distinct from a bounce. AWS’s event schema describes content rejection, including virus detection. Automatically treating every rejection as a nonexistent address can hide the actual problem.
A receiver can accept a message and later return a non-delivery report. Investigate that sequence using the original message ID and diagnostic details. A Transient/General label alone does not identify the receiving provider or prove that this sequence occurred.
Add an application-level review process
The useful second layer is a decision process that connects failure history with the next send:
- Check suppression configuration. Enable the appropriate bounce and complaint reasons and review overrides. Preserve unsubscribes separately from delivery failures.
- Capture useful evidence. Configure SES event publishing and retain message IDs, recipient addresses, event timestamps, classifications, and diagnostic responses. Keep access and retention appropriate for the recipient data you store.
- Stop repeat attempts to known blocked recipients. Check your application’s send eligibility before calling SES. Make sure every sending path uses the same rules.
- Review recurring temporary failures. For example, three transient bounces in 30 days could place an address in a review queue. This is an illustrative operating policy, not an AWS rule or proof of permanent failure.
- Separate address problems from message problems. A malformed attachment needs a content fix. An invalid mailbox needs an address correction. A complaint needs a consent review.
- Record recovery decisions. Require evidence that the underlying problem is resolved before clearing a block, then watch the next delivery attempt.
These controls matter even when reputation metrics look healthy: repeated failures can mean a customer is missing an invoice, a sign-in link, or a product notification.
Where SendOps fits
SendOps connects to your AWS account and combines delivery reporting with contacts, segments, and email workflows. Sending uses your SES account; SendOps holds contacts, consent, events, and campaign content.
Its Undeliverable view groups addresses by failure signals and configurable recurrence rules. This is separate from the AWS SES suppression list. Use the recorded reason and diagnostic to investigate; a row is not, by itself, proof that the mailbox no longer exists. Clearing one list does not clear the other or restore marketing consent.
If you already have a reliable event pipeline and a support workflow, build on them. If your team needs a shared place to investigate failures alongside its email workflows, explore SendOps.
For the next operational step, see how to manage SES suppression safely, what to monitor in SES, and how to investigate bounce and complaint rates.