An Amazon SES Send event means SES will attempt delivery; a Delivery event means the recipient’s mail server accepted the message, not that it reached the inbox. That difference matters when you search “transactional emails not sending”: you need to trace the message from its trigger to the recipient, not reset SMTP settings blindly.

Transactional emails fail at four distinct points: the application never triggers the send, a job waits in a queue, the API or SMTP server rejects submission, or the provider accepts the message but it bounces or is filtered. Start with the provider’s event record and match it to your application log. The first stage with missing or negative evidence is where to investigate.

Transactional emails not sending: locate the failed stage

Start with the provider event or status record. Search by recipient and approximate send time, then follow the message ID if one exists. A record from Amazon SES, Mailgun, SendGrid, or another provider can tell you whether the request arrived and what happened after acceptance.

EvidenceWhat it establishesExample next step
No provider record or message IDThe request may not have reached the providerInspect the application event and job queue
API 401 or SMTP 535 5.7.8Submission failed before message acceptanceCheck authorization or credentials
SMTP 250 after message dataThe SMTP server accepted the transactionFollow the provider’s later delivery events
Bounce or suppression eventThe provider did not complete delivery to that recipientRead the event reason and recipient status
Delivery eventThe receiving mail server accepted the messageCheck inbox, junk, and recipient-side filtering

For SES, a Send event indicates an attempt to deliver; a Delivery event indicates acceptance by the recipient’s mail server. Neither confirms inbox placement. Amazon SES describes its sending events.

A practical example: an order-confirmation job logs order.created, but there is no provider message ID. That points upstream of delivery—perhaps the worker did not run or the send call failed before it reached SES. If SES instead records Send followed by Bounce, the application submitted the message; the recipient outcome is the next thing to diagnose.

Check the trigger, template, and send queue

Verify that the event occurred and that the message should have been sent for that event. For a SaaS password reset, check that the reset action created a job for the correct user, that the job ran, and that it passed the intended template and recipient address. For WooCommerce, confirm that the order reached the status that triggers the relevant notification and that the email is enabled. A successful test send does not prove that a real application event uses the same template, data, or sending path.

Inspect required template values before calling the provider. An absent recipient, order number, or reset URL can cause a rendering or validation failure. Log a safe event name, template identifier, recipient, and job result; do not put passwords, reset tokens, or other secrets in logs.

For WordPress specifically, WP-Cron checks scheduled tasks on page loads rather than running continuously. A low-traffic site can therefore process a due task later than its scheduled time. WordPress’s default intervals include hourly, twice daily, daily, and—since WordPress 5.4—weekly; plugins can add intervals with the cron_schedules filter. If a WooCommerce message is waiting in a scheduled queue, inspect the queue and Tools > Site Health for cron-related errors. This is a WordPress-specific possibility, not a general explanation for a SaaS worker that never ran. WordPress explains WP-Cron scheduling.

Verify API or SMTP connection settings

A connection or authorization error means the message has not cleared submission. Check the configured endpoint or SMTP host, port, encryption mode, credentials, sender address, and request format against the provider’s settings. Amazon SES, Mailgun, and SendGrid have provider-specific endpoints and credentials; a hostname or TLS setting from one service does not carry over to another.

Sending pathConcrete evidenceWhat it means
REST APIHTTP 401 responseThe request was not authorized
SMTP AUTH535 5.7.8Credentials were invalid or insufficient
SMTP AUTH530 5.7.0Authentication is required but is not active
SMTP after DATA250The server accepted the submitted message

For SMTP submission, port 587 is conventionally used with STARTTLS; port 465 uses implicit TLS, where the TLS handshake begins as soon as the connection opens. These are protocol conventions, not universal provider settings. Port 25 is not the reserved submission port, although RFC 6409 allows a service to designate it for submission. Follow the actual provider’s host and encryption configuration rather than changing ports on guesswork (RFC 6409; RFC 8314).

Here is an illustrative SMTP exchange. The replies are standard examples, not a transcript from a particular provider:

C: AUTH PLAIN [credentials]
S: 535 5.7.8 Authentication credentials invalid

That failure happens before message content is submitted. After correcting the credentials, a successful exchange might continue like this:

S: 235 2.7.0 Authentication successful
C: MAIL FROM:<receipts@example.com>
S: 250 2.1.0 OK
C: RCPT TO:<customer@example.net>
S: 250 2.1.5 OK
C: DATA
S: 354 End data with <CR><LF>.<CR><LF>
C: [message headers and body]
C: .
S: 250 2.0.0 Accepted

The client sends the message only after 354; the final 250 says the server accepted responsibility for the transaction, not that the message is in the inbox (RFC 5321; RFC 4954). For Volanea SMTP, the username is volanea and the password is the project’s secret API key. A sk_test_ key renders and logs messages but does not deliver them. The From field accepts a bare address or Name <address>.

For an API, retain the HTTP status, response body, and returned message ID. For example, this illustrative—not provider-specific—error body points to sender validation rather than inbox filtering:

{"error":"invalid_from","message":"From address is not verified"}

Volanea supports a REST API and SMTP relay, so first identify which path your application actually calls. The Email API Send guide covers the API submission flow.

Read the provider response and message status

An API response or SMTP reply describes submission; later events describe the provider’s delivery attempt. In SMTP, 2xx is a positive completion, 3xx an intermediate response, 4xx a transient negative completion, and 5xx a permanent negative completion. A 4xx may be retried depending on the provider and error; a 5xx usually requires correcting the cause before another attempt. Provider event names and retry policies vary.

Use the message ID to connect your application log to provider events. Volanea records sends and provides delivery, open, click, bounce, and unsubscribe statistics, plus event logs and contact activity timelines. That evidence is useful in a specific case: if a request is accepted but a recipient is suppressed, the contact’s activity and suppression state distinguish that from a send request that never arrived.

Interpret bounce labels by their reason, not the label alone. A soft bounce generally describes a temporary delivery failure, such as a receiving system deferring the message; a hard bounce describes a permanent failure, often an invalid destination. Providers can expose different event names and details, so preserve the reason code or text with the status. Volanea automatically suppresses hard bounces and handles bounce and complaint suppression. Retrying a hard-bounced address will not solve the underlying address problem.

A minimal event trace might look like this:

10:14:02 app: password_reset_requested user=42
10:14:02 worker: template=password-reset recipient=user@example.net
10:14:03 provider: accepted message_id=abc123
10:14:08 provider: bounce type=permanent recipient=user@example.net

Here, the trigger and submission succeeded. The bounce—not the template or API connection—is the failure evidence to follow up. If the provider instead records delivery, investigate inbox placement or recipient filtering.

Confirm sender-domain authentication

Check the domain in the visible From address and the identities used by the actual sending service. SPF authorizes hosts for the SMTP HELO or MAIL FROM identity. DKIM adds a signature that the receiving system verifies using a public key in DNS. DMARC evaluates the visible From domain and passes when at least one passing SPF or DKIM identity aligns with it (RFC 7208; DKIM specification; RFC 7489).

Use the DNS records supplied for the sending provider rather than adding records by imitation:

  • SPF: Look up TXT records for the relevant domain. There should be one SPF record beginning v=spf1; multiple SPF records for the same name produce permerror. SPF evaluation is limited to 10 DNS-query-causing terms, including include, a, mx, ptr, exists, and redirect. Exceeding that limit also produces permerror.
  • DKIM: Use the signature’s d= signing domain and s= selector to locate the key at <selector>._domainkey.<signing-domain>. A missing key can cause verification failure.
  • DMARC: Inspect the TXT record at _dmarc.<visible-From-domain> and the message’s Authentication-Results header. The published p policy may be none, quarantine, or reject; the record’s existence alone does not reveal its policy.

Two DNS formatting details can explain hard-to-spot SPF errors: each TXT character-string is limited to 255 octets, but a TXT record may contain multiple strings that SPF processing concatenates without spaces. Also, an mx mechanism that causes more than 10 A or AAAA lookups for an MX record returns permerror. Check DNS and the provider’s domain verification result before editing records. Volanea provides guided SPF, DKIM, and DMARC setup with automatic checks through its Amazon SES setup.

Check the recipient, reputation, and provider limits

If a message is accepted but missing for one person, compare event results by recipient domain. A message reaching Gmail but bouncing or disappearing for a Hotmail recipient is not evidence of a site-wide outage; inspect the Hotmail-specific bounce reason or delivery event, then compare it with a successful recipient. An invalid mailbox, a full or rejecting inbox, suppression, or spam filtering can affect one address or provider differently from another.

Check the suppression list before retrying. A hard bounce or complaint can prevent future sends to that address; in Volanea, hard bounces are automatically suppressed. If a provider reports a rate limit or your application has stopped submitting at a volume boundary, resolve that condition rather than repeatedly resending the same job. On Volanea, sending pauses at the monthly email limit, with no overage billing.

Reputation and network restrictions are separate failure paths. A spam filter can place an accepted message in junk, while a hosting provider or firewall can block outbound SMTP before it reaches the email service. If all SMTP connections from a web host fail, ask the host whether outbound connections to the configured service are permitted. Volanea sends from shared Amazon SES infrastructure and does not offer dedicated IPs; its reputation guard can slow sending before a bad list harms a domain. If a list-quality issue coincides with a sudden increase in bounces, inspect the affected recipients and event reasons instead of treating every delayed message as a credential problem.

Run a controlled test and prevent repeat failures

Send a minimal message to a mailbox you control using the same sender domain and transport as the affected workflow. Record the trigger time, application job result, API response or SMTP reply, message ID, and later provider event. Then trigger the real event and compare the paths. If the test creates a provider record but the real event does not, focus on the real workflow’s trigger, job, or template data. If both are accepted but only one is missing, compare recipient-specific outcomes.

Change one thing at a time and repeat the test. For example, if SPF was corrected, record the DNS change and then inspect a newly sent message’s authentication results; an old message cannot show the effect of a new record. Keep logs useful but safe: retain identifiers and status details, and avoid storing message secrets or unnecessary personal data.

A provider’s acceptance response is one checkpoint, not the final test. Track the later status for a sample of real transactional messages and alert on repeated API errors, SMTP authentication failures, bounces, or suppression growth. Our email deliverability and sending guides cover additional operational checks. Start free with 1,000 emails a month at Volanea.

FAQ

Should I switch from PHP mail to SMTP or an email API?

Switch when the evidence shows PHP’s local mail path is failing or cannot provide the provider response and message ID you need to trace sends. An authenticated SMTP relay or API gives your application a direct submission path to a sending service, but neither repairs a missing trigger, invalid recipient, or message filtered after acceptance. Choose based on the integration and diagnostics your application needs.

Can I send real test emails through Volanea without SES production access?

Yes. After verifying your domain, Volanea lets you send to real recipients without requesting Amazon SES production access. That makes a controlled test possible before migrating a live workflow: verify the domain, send to a mailbox you control, and inspect the resulting send and delivery events.

What happens when I reach Volanea’s monthly sending limit?

Sending pauses at the monthly email limit; Volanea does not bill overages. If transactional messages stop near the limit, compare the account’s usage with the time of the failed request before debugging the template or DNS. The plan features remain available; plans differ by monthly volume and domain count.

Start free with 1,000 emails a month → https://app.volanea.com/signup.