Amazon SES infrastructure is solid - but it doesn't protect you from your own list quality. AWS publishes two complaint thresholds: 0.1% recommended (stay below) and 0.5% (above this your account may be paused) (AWS, Amazon SES FAQ, retrieved 2026-06-07). The infrastructure can't fix a dirty list or a poorly warmed IP. For Mailblast's view of these numbers, see AWS SES bounce and complaint rates.
This guide covers every deliverability lever available to SES senders in 2026: the warm-up process, bounce and complaint handling, reputation monitoring, the new inbox placement metrics, and the list hygiene practices that keep complaint rates in the safe zone.
Key Takeaways
- AWS publishes two thresholds for each rate: 5% bounce / 0.1% complaint (recommended) and 10% bounce / 0.5% complaint (above this, your account may be paused).
- New in 2026: SES now provides inbox placement rates by sending domain and campaign via Virtual Deliverability Manager.
- Hard bounces above 5% indicate a serious list quality problem - suppress those addresses immediately.
- IP warm-up (4–8 weeks for dedicated IPs) is required before sending large volumes from a new IP.
What Amazon SES Manages vs. What You Must Manage
Deliverability responsibility is split between SES and you. Getting this wrong is the single most common reason SES accounts get suspended. For the wider picture see our SPF / DKIM / DMARC primer, the why emails go to spam deep dive, and the help-doc Improving Email Deliverability.
SES handles:
IP pool reputation management on shared pools
DKIM signing (once you configure the CNAME records)
Feedback loops with major ISPs (Gmail, Outlook, Yahoo)
Bounce classification (hard vs. soft) and routing to your SNS endpoint
Blocklist monitoring for your sending domains (with VDM)
Inbox placement data by domain and campaign (with VDM, launched 2026)
You must manage:
List quality: confirmed opt-in, regular purging of inactive addresses
Suppression list maintenance: never re-send to hard-bounce or complaint addresses
Sending volume consistency: erratic bursts after long dormant periods damage reputation
Content quality: spam filter scoring depends on your copy, subject lines, and links
IP warm-up: mandatory for new dedicated IPs, recommended for new shared sending patterns
Understanding SES Sending Reputation Metrics
In 2026, Amazon SES provides three core reputation signals in the SES console.
Bounce rate. The percentage of sends that result in a bounce. Hard bounces (permanent failures: invalid address, domain doesn't exist) count heavily against your reputation. AWS's published thresholds are 5% recommended (stay below) and 10% (above this your account may be paused), measured on a representative sample of recent sending rather than your all-time history (AWS, Managing Amazon SES Sending Limits, retrieved 2026-06-07).
Complaint rate. The percentage of recipients who mark your email as spam. AWS publishes two thresholds here too: 0.1% recommended and 0.5% (above which sending can be paused) (AWS SES FAQ, retrieved 2026-06-07). On a list of 10,000, that's about 10 complaints per send before you cross the recommended line.
Inbox placement rate (new in 2026). Via Virtual Deliverability Manager, you can now see what percentage of your emails land in the inbox vs. the spam folder at major mailbox providers before and after sending. This is a significant improvement over the previous delivery-only visibility.
IP Warm-Up: Why It Matters and How to Do It
Sending a large volume of email from a new IP address - even on Amazon SES - will hurt your deliverability. Mailbox providers track IP history. A new IP with no sending record that suddenly sends 100,000 emails looks like a spam operation.
For shared IP pools: AWS manages warm-up automatically. Shared pools have established sending history, and SES routes your traffic through pools appropriate for your volume. No action required from you.
For standard dedicated IPs ($24.95/month): You are fully responsible for the warm-up schedule. A typical ramp for a 100K/month sending volume:
| Week | Daily Send Volume |
|---|---|
| 1 | 200–500 |
| 2 | 1,000–2,000 |
| 3 | 5,000–10,000 |
| 4 | 20,000–30,000 |
| 5–6 | 50,000–75,000 |
| 7–8 | Full volume |
Start with your most engaged subscribers (recent openers, recent clickers). These generate positive engagement signals that accelerate reputation building.
For managed dedicated IPs ($15/month + usage): AWS handles warm-up automatically, scaling the IP pool as your volume grows.
For a full week-by-week ramp see the SES warm-up guide; for whether a dedicated IP is the right choice in the first place see Amazon SES dedicated IP.
Bounce Handling: Setting Up SNS Notifications
Amazon SES can route bounce and complaint notifications to an Amazon SNS topic, which then delivers them to an HTTP endpoint you control (or your marketing frontend). This pipeline is what allows your sending system to automatically suppress future sends to problematic addresses.
Without this setup, you're flying blind - SES delivers the notifications but you have no automated way to act on them.
The configuration in the SES console:
Create an SNS topic for bounces
Create an SNS topic for complaints
In SES, set up a Configuration Set that routes bounce/complaint events to those topics
Subscribe your endpoint (or your marketing frontend's webhook URL) to the SNS topics
Verify that bounce events trigger suppression in your list
Most SES frontends handle this configuration on your behalf. With Mailblast specifically, the IAM access key you paste in includes a few SNS permissions so Mailblast can create the topics it needs and subscribe itself - there's no webhook URL to copy around. See the bounce-handling deep dive.
List Hygiene: The Highest-Leverage Deliverability Action
The single most effective deliverability improvement available to most senders isn't infrastructure - it's list quality. In 2026, email addresses decay at roughly 22% per year: people change jobs, abandon email addresses, or simply stop engaging (AWS SES, Deliverability Best Practices documentation, retrieved 2026-06-07).
Confirmed double opt-in. Requires new subscribers to verify their email address before being added to your list. Eliminates typos, fake signups, and addresses entered without permission. AWS strongly recommends confirmed opt-in for all marketing email via SES.
Engagement-based suppression. Identify subscribers who haven't opened or clicked anything in 6–12 months. Send a re-engagement sequence. Remove those who don't respond. Sending to a large segment of disengaged subscribers is one of the fastest ways to accumulate spam complaints.
Hard bounce suppression. Every hard-bounce address must be permanently removed from all future sends. SES maintains an account-level suppression list - addresses that hard-bounced in the past are automatically suppressed by SES even if you try to send to them again. Your frontend tool should maintain its own parallel suppression list.
Regular cleaning cadence. Audit inactive subscribers quarterly. Import suppression lists from SES's account-level list to your frontend to ensure they stay in sync.
Virtual Deliverability Manager: Is It Worth $0.07/1,000 Emails?
In 2026, Amazon SES's Virtual Deliverability Manager (VDM) adds inbox placement rates, content testing, and blocklist monitoring for $0.07 per 1,000 emails - a 70% premium on top of the base sending cost.
When VDM is worth it:
You send high-stakes marketing campaigns where inbox rate directly affects revenue
You're troubleshooting a deliverability problem and need visibility into where emails are landing
You want to test new email content before sending to your full list
When VDM isn't worth it:
Your complaint and bounce rates are in the safe zone and you're not actively troubleshooting
Your sending volume is low enough that the additional cost isn't material to include insights
You're using a frontend tool that provides its own engagement analytics (open rates from tracked sends are a proxy for inbox placement)
For most senders under 100,000 emails/month with clean lists, standard SES monitoring is sufficient. Enable VDM during active troubleshooting or campaign optimization periods, then disable when not needed.
Frequently Asked Questions
What is a good bounce rate for Amazon SES?
Amazon SES expects hard bounce rates below 5% for normal operation. Best practice is to keep hard bounces below 2% - above 5%, AWS may reduce your sending limits. If your bounce rate exceeds 10%, stop sending immediately and clean your list before resuming (AWS SES documentation, retrieved 2026-06-07).
What complaint rate will get my Amazon SES account suspended?
AWS publishes two thresholds for complaints: 0.1% recommended (stay below) and 0.5% (above this, your account may be paused). On a list of 10,000, the 0.1% ceiling works out to about 10 spam complaints per send. Consistent double opt-in practices keep complaint rates well below this threshold for most legitimate senders.
Does Amazon SES have spam filtering for outbound email?
Amazon SES doesn't filter your outbound email for spam content - that's your responsibility. AWS does monitor aggregate metrics (bounce rates, complaint rates) and will act on accounts that generate excessive negative signals. Use inbox preview tools and test sends to check spam scoring before sending to your full list.
How do I check my Amazon SES sender reputation?
In the SES console under "Reputation metrics," you'll see your bounce and complaint rates. With VDM enabled, you'll also see inbox placement rates by domain and campaign. AWS also sends email notifications when your reputation metrics approach thresholds that require attention.