Email Marketing on Amazon SES: What Teams Get Wrong (and How to Fix It)
Learn the 5 most common mistakes technical teams make when running email marketing on Amazon SES - from mixed sending domains to bad bounce management.
Amazon Simple Email Service (SES) is an incredibly reliable, low-cost delivery engine. It is the infrastructure of choice for teams that value cost-efficiency, technical control, and ownership of their sending reputation. However, because Amazon SES stops at raw delivery, it leaves the application layer entirely up to you.
When teams transition from sending basic transactional emails to running full amazon ses email marketing campaigns, they often assume they can use the exact same pipeline. This assumption is where things break down. Without a proper platform layer on top of your mailer services, simple operational gaps can quickly ruin your deliverability, trigger AWS account pauses, or violate compliance regulations.
Here are the five most common mistakes technical teams make when running ses marketing campaigns, and how to architect a proper solution.
Mistake 1: Not Separating Transactional and Marketing Email
Many teams use a single AWS account, a single verified domain, and a single set of credentials to send both critical signup tokens and weekly marketing newsletters. This is a severe reputation risk.
Transactional emails (such as password resets, invoices, and shipping alerts) have near-zero complaint rates and high open rates. Marketing emails (such as product updates, digests, and newsletters) naturally draw higher bounce and complaint rates.
If you mix these flows on the same sending identity, a spike in spam complaints on a marketing broadcast can damage the IP or domain reputation for your entire system. In worst-case scenarios, AWS will pause your SES sending capabilities entirely, meaning your application can no longer deliver critical OTP tokens or password resets.
How to fix it:
- Verify distinct sending identities: Use subdomains to separate your flows. For example, use
mail.yourdomain.comfor transactional messages andnews.yourdomain.comfor marketing broadcasts. - Leverage Configuration Sets: Map your different email types to dedicated SES Configuration Sets. This allows you to apply different tracking, IP pools, or rules depending on whether the message is transactional or promotional. Review SES best practices for deeper architectural strategies.
Mistake 2: Missing or Broken Unsubscribe Handling
When a recipient wants to stop receiving your marketing emails, they need a clear path to do so. Under major inbox provider guidelines from Google and Yahoo, bulk senders must support a clear unsubscribe link as well as RFC-defined multipurpose internet mail extension headers for one-click unsubscribe handling.
SES does not automatically parse your email body to inject a working unsubscribe footer or track who has opted out. It simply routes the bytes you provide. If your system does not handle opt-outs, frustrated users will hit the “Mark as Spam” button. This directly increases your complaint rate, which AWS monitors closely.
How to fix it:
- Include a standard
List-UnsubscribeandList-Unsubscribe-Postheader in all marketing sends. - Maintain a centralized suppression list at the application level. Before sending any marketing broadcast, your system must cross-reference this database to ensure no opted-out addresses receive the mail.
Mistake 3: Treating All Bounces the Same
When an email delivery fails, SES generates a notification containing details about the bounce. However, many teams make the mistake of handling all bounces under a single, aggressive rule: either they ignore them entirely (risking AWS suspension), or they immediately block the recipient forever.
This lack of nuance causes major issues:
- Hard Bounces: The email address does not exist or has been permanently blocked. Sending to these repeatedly will ruin your reputation. They must be removed immediately.
- Soft Bounces: The recipient’s mailbox is temporarily full, or their server is experiencing a transient error.
Treating a temporary soft bounce as a permanent failure means you accidentally suppress valid customers. Conversely, ignoring hard bounces leads directly to your account being placed under review by AWS.
How to fix it:
- Parse the bounce type from your SNS notifications or SES event destinations.
- Automatically suppress hard bounces on the first occurrence.
- Implement an incremental backoff strategy for soft bounces, only suppressing the contact if the delivery fails consistently over several days.
Mistake 4: Zero Visibility Into Your Sending History
When a customer support agent asks “did we send the invoice to user@example.com yesterday?” or “why did this customer stop getting our product updates?”, standard AWS tooling falls short.
CloudWatch metrics can show you a high-level aggregate graph of your bounce rate, but they cannot tell you which specific message bounced, who triggered it, or what the template rendered. Finding a specific message requires digging through CloudWatch logs, which is time-consuming and inaccessible to non-technical team members.
How to fix it:
- Store a message-level delivery log that maps each unique SES message ID to the recipient, the campaign, and its final delivery status (Delivered, Bounced, Complained).
- Build a simple interface or adopt a platform layer that allows your support and marketing teams to search sending history by recipient email address without requiring AWS Console access.
Mistake 5: No Alerting Until the Account is Suspended
AWS maintains strict thresholds for SES accounts: your bounce rate should remain below 2% (and must stay under 10%), while your complaint rate must stay below 0.1% (and must stay under 0.2%).
If you exceed these limits, AWS does not just send a warning. They can place your account under review or suspend your sending capabilities entirely. Relying on manually checking your AWS dashboard once a week to monitor these metrics is a recipe for sudden downtime.
How to fix it:
- Set up proactive CloudWatch Alarms on your reputation metrics.
- Configure alerts to trigger when your bounce rate crosses 2% or your complaint rate hits 0.05%, giving your engineering team time to pause campaigns and investigate before AWS takes action.
Building vs. Adopting a Platform Layer
To run email marketing aws setups safely, your team is essentially forced to build a custom application on top of SES. You have to write subscription management code, configure SNS topics, run database migrations for suppressions, build a campaign builder, and write custom monitoring scripts.
This is why we built SendOps. SendOps connects directly to your existing AWS account using secure, short-lived credentials. It leaves your production sending path completely untouched while instantly adding the entire marketing and operations layer: dynamic audience segmentation, automated workflows, reliable unsubscribe handling, unified analytics, and proactive deliverability alerting.
Instead of spending engineering hours building and maintaining email glue-code, you get to keep the low cost and reliability of Amazon SES while giving your entire team a modern, production-grade email marketing platform.