Help Monitoring
Alerts and monitoring
This page explains how to monitor your sending in SendForever: the numbers on the team Overview and on a domain's Overview, the bounce and complaint levels Amazon SES judges you by, the alerts your AWS account sends when something goes wrong, who gets them by email, and the Alerts page.
The team Overview
The app opens on the team Overview. Besides the team's name and plan, it has three parts.
Sending shows all domains together over the last 30 days, in four cards: Sent (accepted by Amazon SES), Delivered, Bounced and Complaints, each with its share of what was sent. The numbers are the totals that Amazon SES publishes on its own to CloudWatch in your AWS account. They cover the whole Amazon SES account in its region. SendForever reads them when you open the page and keeps them for up to 15 minutes. If CloudWatch cannot be read, the page says Unable to read your sending numbers and shows a Try again button.
Each card has a badge that compares the period with the 30 days before it, up to yesterday. For Sent and Delivered it compares the average per day. For Bounced and Complaints it compares the rate of what was sent. Green means good news and red means bad news, so sending more at the same bounce rate does not show as bad. Bounced and Complaints also carry the reputation badge described below.
Issues lists what is still open. Each row leads to where you resolve it:
- Alerts that Amazon SES reported and has not cleared yet. These open the Alerts page.
- Domains with a check that fails, waits on the DNS or needs attention. These open the domain's settings.
- The sandbox, while your Amazon SES account only delivers to verified addresses, up to 200 emails a day. This opens the AWS account page. See Production access.
When nothing is open, the Overview leaves Issues out.
Recent domains lists the five most recently changed domains, with a link to all of them.
A domain's Overview
A domain's Overview shows how email from the domain chosen in the header did over the last 30 days:
- Needs attention: the alerts still open about this domain and about the whole account.
- Sent, Delivered, Bounced and Complaints: the domain's own totals, with the same comparison badges as the team Overview. With deliverability insights on, the totals are what Amazon SES's Virtual Deliverability Manager counted for the domain's configuration set, in whole days up to yesterday. A day the insights did not count in full, such as one before they were turned on, comes from the activity log. Without insights, they are counted per recipient from that set's emails in the activity log, today included. If the domain still uses a set that other domains share, the insights count it by its name, because the set also carries their emails. If insights cannot be read, the numbers come from the log instead.
- Opened and Clicked, for a domain that tracks opens and clicks: emails opened or clicked at least once, as a share of what was delivered since the first open or click the log has in the period. Each email counts once, on the day it was sent. A domain that does not track them says so and shows where to turn tracking on.
- Emails per day: a chart of sent, delivered, bounced and complained-about emails per day. When the numbers come from the activity log, today is labeled as so far. When they come from the insights, the chart ends yesterday. A domain that sent nothing says so instead of showing an empty chart.
The domain's numbers need the activity log. If the log cannot be read, the page says Unable to read the activity log with Try again.
The domain's Settings page also has a Reputation tab. It shows the domain's bounce and complaint rates and the levels to keep them under. Once you turn on deliverability insights (Amazon SES's Virtual Deliverability Manager) on the AWS account page, it also shows the figures they keep for the domain. See Your AWS account.
Reputation levels
Amazon SES judges an account by its bounce rate and complaint rate, which it publishes to CloudWatch. SendForever shows the latest rate in a badge:
| Badge | Bounce rate | Complaint rate |
|---|---|---|
| Healthy | under 2% | under 0.05% |
| Watch it | from 2% | from 0.05% |
| Too high | from 5% | from 0.1% |
Amazon SES places an account under review at a 5% bounce rate or a 0.1% complaint rate, and pauses sending at 10% and 0.5%.
On the team Overview, the badge shows the whole account's rate. On a domain's Overview, it shows the rate of the domain's configuration set once Amazon SES measures that set separately, and the account's rate until then. Hover over or focus the badge to see whose rate it is, for example "Domain rate: 0.40%, SES reviews at 5%".
The alerts your AWS account sends
The SendForever stack creates the alerts in your account:
- An Amazon SNS topic named
sendforever-alerts. - Two CloudWatch alarms that publish to it.
sendforever-bounce-ratepublishes when the account's bounce rate, averaged over an hour, reaches 5%.sendforever-complaint-ratepublishes when its complaint rate reaches 0.1%. Each alarm also publishes when the rate is back under its level. - Two EventBridge rules that publish to it.
sendforever-tenant-statuspublishes when Amazon SES pauses or resumes an isolated domain.sendforever-tenant-findingspublishes when Amazon SES finds or clears a problem with one.
SendForever subscribes its own endpoint to the topic. It only trusts messages that Amazon signed and that come from the topic in your account. Anything else in your account can subscribe to the topic too.
The alerts you can get:
| Alert | When |
|---|---|
| Bounce rate too high: 5.2% | The account's bounce rate reached 5% |
| Bounce rate back under 5% | The rate came back under 5% after being too high |
| Complaint rate too high: 0.12% | The account's complaint rate reached 0.1% |
| Complaint rate back under 0.1% | The rate came back under 0.1% after being too high |
| Sending paused for example.com | Amazon SES paused an isolated domain because of its reputation, or someone paused it by hand from SendForever or the AWS console |
| Sending resumed for example.com | The domain sends again |
| Amazon SES found a problem with example.com: bounce rate (high impact) | Amazon SES found a problem, such as DKIM, DMARC, SPF, BIMI, bounce rate, complaint rate, feedback loop or IP blocklisting |
| The bounce rate problem Amazon SES found with example.com is resolved | Amazon SES cleared it |
An alarm's first OK only means the alarm has data, so it is not an alert. Amazon SES repeats a finding while it lasts. SendForever keeps only the first report, and keeps a resolution only if it recorded the problem.
The domain's Automatic pausing setting, on its Configuration tab, decides when Amazon SES pauses an isolated domain on its own. See Domains.
Campaign stops
A campaign stops itself when too many of its emails bounce or draw complaints. The team's owners and admins then get an email with a subject like "[Acme] Spring sale stopped". It gives the rates at the moment the campaign stopped and has a Review the campaign button. A campaign stop is reported by email and on the campaign's page, not on the Alerts page. See Campaigns.
Who is told
Every alert is emailed to the team's owners and admins as it arrives, except a low impact finding about the whole AWS account, which is only listed on the Alerts page. Members do not get alert emails. The email's subject is the team's name in brackets followed by the alert, such as "[Acme] Bounce rate too high: 5.2%". The email says when Amazon SES reported the alert (in UTC) and what to know about it. For a DKIM, DMARC, SPF or BIMI finding, it also says what to do. It has a See the alert button.
The AWS account page lists who gets alert emails under Alerts are emailed to:. To add someone, invite them as an admin. This needs the Company plan. See Teams.
The Alerts page
Open the team's Alerts to see what Amazon SES reported about your account and domains.
The bell in the header, on every page, shows in red how many problems are open. These are the same problems Needs attention lists. Click the bell to see the latest five and when each started. Click one to open it on the Alerts page, or click All alerts to open the page itself. When nothing is open, the bell shows no number and says so. Scripts and AI agents read the same problems with list_alerts (see API and MCP).
Needs attention lists the problems that are still open, each collapsed to its title. Open one to see its details. For a bounce or complaint rate, a paused domain, or a problem Amazon SES found with a domain's DKIM, DMARC, SPF, BIMI, bounces, complaints or feedback loop, it also shows what to look at:
- What to do: for example, the domain most bounces or complaints come from, with a button to open it. It can also say that the suppression list is off, with Turn it on, or that you send much more per day than in the previous 30 days. For a DKIM, DMARC or SPF finding, a button opens the domain's DNS records tab, which shows what to publish or change. For a bounce, complaint or feedback loop finding, See reputation opens the domain's Reputation tab with its rates at each mailbox provider. For a paused domain, Open configuration opens the tab where you resume sending.
- Already done: what is already in place, such as the suppression list being on.
- Good to know: for example, Gmail does not report complaints to Amazon SES, so they are not in the complaint rate (Google Postmaster Tools shows them). And pausing does not bring the rate down, because Amazon SES measures your recent sending.
A BIMI finding means Amazon SES suggests a brand logo for the domain. Some inboxes show this logo beside the domain's emails. The alert says what the domain still needs. Set up brand logo opens the Brand logo (BIMI) card on the domain's DNS records tab. Upload a square SVG Tiny PS file under 32 KB there, and SendForever hosts it in your AWS account and adds its BIMI record. A logo only shows once DMARC is at quarantine or reject. Until then, the alert asks you to change DMARC first, with Change DMARC. For a domain SendForever has not checked yet, the alert asks you to check its records first, with Check DNS records. A brand logo is optional: email is delivered the same without one. See Domains.
Last 90 days lists every incident from the last 90 days, newest first. Each one shows a red dot while it is open and a green one once resolved, when it started and how long it lasted. Click an incident about one of your domains to open the tab where you take its next step. An open incident names that step, such as "Next: Set up brand logo" or "Next: See reputation". An incident with no step left, such as a resolved one, opens the domain's Overview.
Alerts cover the domains this team has in SendForever and the AWS account as a whole. Amazon SES also reports on other domains in the account, such as one you moved to another team or never added. Those are not listed or emailed here. A domain you add shows its findings the next time Amazon SES reports them. If nothing was reported, the page says Amazon SES reported nothing in the last 90 days. Alerts are kept for 90 days.
The See the alert link in an email opens this page with that alert open, or marked in the list if it is already resolved.
A problem stays open until Amazon SES reports that it has cleared. Amazon SES offers no way to close a finding about a domain, and a low impact finding can stay open for weeks. So owners and admins can click Ignore on an open finding. The finding then leaves Needs attention and the Overview. It stays in Last 90 days with an amber dot and the name of who ignored it, and Undo there brings it back. An ignored finding comes back on its own if Amazon SES raises its impact, or clears it and reports it again. Either way, you get an email as for a new finding. A bounce or complaint rate and a paused domain cannot be ignored, because they are what stops your sending.
When alerts stop arriving
If your AWS account is no longer sending its alerts to SendForever, the Alerts page and the domain's Overview show Alerts are not reaching SendForever. Until they are connected again, nothing new appears on the page or by email. Owners and admins can click Connect alerts to go to the AWS account page. There, the Alerts card shows Connected, Connecting (while AWS confirms the link, within a minute) or Not connected with Connect alerts. The card also shows each alarm's state: fine, too high or no data yet.
Common questions
I got "Bounce rate too high". What should I do?
Open the alert on the Alerts page. It names the domain most bounces come from, checks that the suppression list is on, and flags a sudden rise in volume. Send only to people who signed up, and stop sending to addresses that bounce.
I got "Amazon SES found a problem with example.com: BIMI". What should I do?
Amazon SES suggests a brand logo for the domain. Open the alert and click Set up brand logo, then upload a square SVG Tiny PS file under 32 KB on the domain's DNS records tab. If the alert asks, move DMARC to quarantine or reject first. Email is delivered the same without a logo. The alert clears once Amazon SES finds the logo's record.
I got "Amazon SES found a problem with the account: engagement". What should I do?
Amazon SES sends this finding about the whole AWS account to suggest tenants, which keep each domain's reputation apart. To put a domain in a tenant of its own, click Isolate this domain in its Settings. If your domains are already isolated, or you choose not to isolate them, click Ignore. This finding is low impact, so it is listed on the Alerts page without an email.
How do I make an alert go away?
Fix the cause. A rate problem resolves when the alarm reports the rate back under its level, and a finding when Amazon SES clears it. Owners and admins can also click Ignore on an open finding to take it off Needs attention and the Overview. A bounce or complaint rate and a paused domain cannot be ignored.
Why don't members get alert emails?
Alerts go to the team's owners and admins. Invite someone as an admin to add them.
Why do the team Overview and a domain's Overview show different totals?
The team Overview reads CloudWatch, which counts the whole Amazon SES account in its region. A domain's Overview counts that domain's emails in the activity log.
Why is my domain's badge different from the account's?
Once Amazon SES measures the domain's configuration set apart, the domain's badge shows that set's rate. The alarms and the team Overview watch the whole account's rate.
Why are complaints from Gmail missing?
Gmail does not report complaints to Amazon SES, so they are not in the complaint rate. Google Postmaster Tools shows them.
Will pausing sending bring my rate down?
No. Amazon SES measures your recent sending, so keep sending email people want.
Why are my Overview numbers a few minutes old?
The team Overview's numbers are kept for up to 15 minutes after they are read from CloudWatch.
The page says "Alerts are not reaching SendForever". What happened?
The subscription of SendForever's endpoint to the sendforever-alerts topic is missing. An owner or admin can click Connect alerts on the AWS account page.
Still stuck?