Emails are not arriving: a marketer's deliverability diagnosis
Separate server rejection from spam placement, review authentication and list quality, define stop conditions, and give the technical team a usable diagnosis.
1. Describe a testable symptom
The phrase our email is broken covers several different failures. A message may remain in the sending queue, be rejected by the receiving server, reach a spam folder, or arrive without prompting a visible action. Start with a specific recipient, timestamp, and message identifier. Inspect the provider status and receiving-server response. Do not ask a customer to share a mailbox password.
Delivered usually indicates server acceptance, not confirmed inbox placement. An open does not prove reading, because mail clients handle image loading differently. Build an evidence trail for a particular message. If complaints cluster around one receiving domain, isolate it. If the entire campaign is affected, investigate changes in configuration, volume, or list composition.
2. Check authentication and provider requirements
Record the From domain, sending service, and SPF, DKIM, and DMARC results from a test message's headers. Give the technical owner these details and any rejection code. Do not edit DNS by guesswork: the same domain may send invoices, CRM notifications, and support messages through different services. Fixing one sender should not break another legitimate workflow.
Gmail distinguishes general sender requirements from requirements for bulk messages to personal Gmail accounts. Its bulk requirements include SPF, DKIM, DMARC, and unsubscribe provisions for marketing messages. Check the current documentation against your sending pattern. Authentication addresses one technical layer; passing it is not a promise that every recipient will see the message in the inbox.
3. Review the audience before resending
Separate the list by origin and recipient expectations: new subscribers, active customers, inactive contacts, and records with an unclear source. An old spreadsheet is not an appropriate audience merely because it was imported into the CRM. Verify why recipients should expect this particular message and where the relevant preference was recorded. Suppress unsubscribed addresses and permanent failures from further sends.
Do not begin recovery with a large campaign to the least engaged contacts. First investigate complaints and errors. Hold segments with unknown provenance until they can be reviewed. The content and frequency should match the subscription promise. Downloading one worksheet does not create an expectation of daily promotions. Legal permission and technical deliverability require separate checks rather than assumptions that one proves the other.
4. Compare segments with consistent denominators
Illustrative example: a provider attempts 1,000 messages, fifty receive permanent failures, and receiving servers accept 950. The permanent-failure rate against attempted sends is 5%. Treat that as a reason to investigate this example, not a universal provider threshold. If nineteen accepted messages produce a tracked click, the observed click rate against accepted messages is 2%.
Do not compare that figure directly with a dashboard dividing clicks by opens or counting repeated clicks. Record the denominator and deduplication method. Break results down by receiving domain, subscription source, and audience age. An overall average can hide one problematic segment. Low click activity alone does not prove poor deliverability: recipients may have received an uninteresting message successfully.
| Illustrative send | Count |
|---|---|
| Attempted | 1,000 |
| Permanent failures | 50 |
| Server-accepted | 950 |
| Messages with click | 19 |
| Click rate against accepted | 2% |
5. Run a controlled diagnostic send
After correcting the suspected cause, test with controlled inboxes and a limited audience that expects the message. Size depends on normal volume and risk; there is no universal warm-up count that fits every sender. Record timestamp, sender, subject, segment, and configuration. Change one meaningful cause at a time so the result remains interpretable. Test the unsubscribe path as well as receipt.
Agree on stop conditions before sending: a new rejection pattern, a material complaint increase, broken unsubscribe processing, or messages reaching suppressed contacts. If a condition occurs, halt subsequent steps and inspect the logs. A second message saying you missed this does not repair a technical failure. Provider support needs message identifiers and a time range, not just a screenshot of aggregate statistics.
6. Add a pre-send decision record
The pre-send record should identify the audience source, exclusions, sender, technical checks, frequency, and the person authorized to pause the sequence. Preserve failure reasons and list changes after each campaign. Maintain suppression consistently across campaigns so a person who unsubscribed does not reappear through another import. Test the complete path from subscription to stopping messages.
The email sequence tool drafts subjects, content, intervals, and exit conditions. It does not inspect DNS, domain reputation, or actual delivery. Use it after the diagnosis establishes the audience and next action. The downloadable worksheet supports a handoff between marketing and the technical owner. A concise evidence record is more useful than speculation that one ordinary word automatically puts every campaign into spam.
Email delivery diagnostic sheet
Document the observation, evidence, and owner before a resend.
Download the working CSV template
Open in Excel or Google Sheets. Replace example rows and empty fields with your own data.
Frequently asked questions
Will changing the subject line fix delivery?
It may affect audience response, but diagnose the cause first. Rejections, authentication failures, and messages sent to suppressed contacts require operational changes.
Does delivered mean inbox placement?
Usually it means server acceptance. Check placement in controlled inboxes, while recognizing that one test address does not represent every recipient's experience.