Missing customer email troubleshooting starts with a time window, the exact address they used, and hop history—not with an apology and a DNS flap. If MailerZ has a hop, you have a code. If history is empty, check leftover MX and whether they sent. HOLD may have the copy. Destination junk is not “they emailed and we never got it” until they search all mail. No inboxing percentage. The store window is the clock.
Quick answer for missing customer email troubleshooting
Customers are not liars and they are not always precise. Evidence splits those.
RFC 5321 replies are the language. Copy them.
Empty history plus leftover MX means they may have mailed the old host.
HOLD means they mailed an unpublished local-part.
250 plus missing UI is search and junk.
Start free so you have history before the first angry ticket.
Authoritative mail transport is defined in IETF RFC 5321 — Simple Mail Transfer Protocol. Product path: delivery recovery, troubleshooting, and docs.
User problem and decision criteria
Decision criteria: window inside store, exact local-part, exclusive MX, destination owner searched.
Do not argue tone. Ask for sent-item proof.
Agencies should not flap MX to soothe a call.
If they mailed a personal Gmail printed on an old PDF, that is not your domain.
No SLA invention.
No SOC 2 as a substitute for history.
Self-send by the customer to themselves is not this ticket.
Portals that cannot forward are a sender class.
Technical mail flow
Their send → their provider → public MX → MailerZ or leftover → HOLD/forward/reject → destination.
You only see what reached you.
Recovery applies to stored hops.
Message-ID if they have one is gold.
Step-by-step setup / decision path
- Collect UTC window, To:, From:, Message-ID if any.
- Public MX two resolvers.
- History lookup.
- Classify.
- If HOLD, map or explain.
- If leftover, tell the truth and delete leftovers.
- If 250, watch them search junk.
- If expired store, ask for a resend to a named alias.
Classify the next failure before a second DNS edit.
HOLD unknown unless you wrote a FORWARD reason.
Quote live pricing before promising alias counts.
Failure modes and proof
DNS flap first.
No time window.
Wrong local-part assumed.
Empty history ignored leftover MX.
Expired store guessing.
Self-send by you as proof.
Promised recover of old-host mail.
Inboxing argument.
Catch-all FORWARD to “not miss again.”
Header rewrite.
Yelling in the ticket.
Closing without a class sentence.
MailerZ workflow and product boundary
MailerZ is custom-domain aliasing and forwarding with optional paid send-as. Secuno LLC operates mailerz.net. The app is mail.mailerz.net. Not Workspace, not IMAP, not an open relay, not a campaign ESP.
Envelope SRS only. Header From, Subject, Date, Message-ID, body, and MIME stay intact. Exclusive MX. Hold unknown on Free. Copy SMTP host, port, and TLS or STARTTLS from the dashboard when you send.
Free: one domain, ten aliases, one seat, fourteen-day store, send-as disabled, SMTP and API disabled. Solo forty dollars a year, twenty-five aliases, ninety-day store, 2,500 outgoing, 20 send-as per hour. Starter eight or eighty. Business nineteen or one hundred ninety. Agency thirty-nine or three hundred ninety. Quote pricing. No SOC 2, ISO, HIPAA, SLA, or inboxing percentage.
Cost, alternatives, and trade-offs
A flaps-first culture costs real mail.
History is cheaper than refunds.
HOLD review finds the typo.
Expired store is why you ask for a resend fast.
Catch-all as a make-good costs junk for months.
Agencies: evidence packets reduce drama.
No fake recoveries.
No inbox KPI.
Operational depth
A short form for customers: time, To, From, sent screenshot. That form is the product of this article.
Support should paste the SMTP code into the ticket, not “looks fine.”
If the customer used a phone share-sheet, they may have mailed a contact card, not the alias. Compare characters.
Shared inboxes: make the owner search on a screenshare.
Store: fourteen Free, ninety paid. Say the date.
Do not recover into Slack. Use the product.
If they never sent, history stays empty and MX is exclusive. Say that kindly with the sent-folder ask.
Portal-only failures: offer an ATS or a form.
No invented traces from Gmail you cannot see—ask them for original if they received a bounce.
Agencies: this is the missing-mail runbook with a customer-facing script.
Re-probe after leftover delete.
Write the class in the reply.
Evidence order that names the hop
“Customer says they emailed us” is not a ticket until you have three artifacts: the exact To: local-part, a time in UTC, and their sent-folder proof or a copy. Without those you are guessing leftover MX, HOLD, destination spam, or a story. MailerZ history is the next artifact. Empty, HOLD, accept 550, or 250 then destination 4xx/5xx. Copy the SMTP line before anyone edits DNS.
Empty history with leftover Google or Microsoft MX means their message never reached MailerZ. Two public resolvers. Delete leftovers. Do not enable catch-all. Do not rotate DKIM. The customer’s Gmail-to-Gmail test may have “worked” for them and still never queried you.
Empty history with exclusive MX means the name never arrived or you probed wrong. Self-send is invalid. Ask them to send a unique subject from the same account they claim they used. If history stays empty, they did not send to your MX or they sent to a typo that HOLD’d on Free.
HOLD is a missing named alias or an unknown on Free. Create the printed local-part. Catch-all FORWARD is paid and optional and is not a debugger. If you FORWARD to “catch the customer,” you hide the next miss.
Accept 550 is MailerZ recipient policy. The name is not allowed. Creating it or telling the customer the correct string is the fix. Republishing MX will not create an alias.
250 then Gmail or Outlook 5xx is destination refuse. The hop worked. Work the destination. Copy the line into the ticket so the next person does not restore aspmx.
250 and the message sits in spam is still a pass for the hop. Open original. Header From should still be the customer. Envelope may show SRS. If someone rewrote Header From, you are not on MailerZ’s hop — we do not rewrite it.
Retention: 14 days Free, 90 paid. If they emailed last month and you are on Free, hop evidence may be gone. The destination inbox is the archive. Say that early. Do not invent a year of store.
Leftover MX plus a proud “we use MailerZ” in the footer is the usual Friday. History looks flaky because half of senders still hit Google. Exclusive means exclusive.
Send-as 550 is unrelated. Do not debug inbound by rotating SMTP. Free cannot send. Unauthorized From is 550 5.7.1.
Agencies: one customer domain per ticket. A blended “three clients say mail is down” thread will get a leftover MX restore on the wrong zone.
Ask for a screenshot of their To: field, not only a story. People send to the old registrar address, a plus tag you never created, or a Google Group that still exists on leftover MX.
After the class is named, do one change. Then probe from another mailbox yourself. If you cannot reproduce, you do not have a product outage. You have a single sender path you have not seen.
No inbox SLA in the reply to the customer. You can show a hop row or an empty row. You cannot promise Primary. You cannot quote review counts. You cannot invent SOC 2 to calm them.
Reply templates that name the class
Reply leftover: “Public MX still lists Google/Microsoft. Their message would not create a MailerZ hop. We deleted leftovers and asked them to resend a unique subject from the same account.”
Reply HOLD: “The To: local-part was not a named alias. Free holds unknowns. We created the name. Please resend.”
Reply accept 550: “The hop refused the recipient. That is policy, not Gmail. We created the alias or we confirmed they used the wrong string.”
Reply destination 5xx: “MailerZ accepted (250) and Gmail/Outlook refused. SMTP line attached. We will not republish MX.”
Reply spam folder: “Accepted and delivered. Open original. Header From should be the customer. If they cannot find it, they should search the unique subject.”
Reply no evidence: “We need UTC time, exact To:, and a sent screenshot. Without those we will not change DNS.”
Reply self-send: “Please send from another provider. Gmail-to-Gmail can hide the hop.”
Do not reply with an inbox percentage. Do not reply with a certification. Do not reply with a review count. Attach the hop row or the empty row.
Internal note: one change after the class is named. If two people edit MX and SMTP in the same hour, you lose causality. The next customer ticket will be worse.
If the customer used a Google Group or a leftover registrar address, say so. Those strings are not MailerZ aliases until you create them and MX is exclusive.
Retention honesty: if they mailed outside 14 days on Free, say the hop row is gone. Search the destination. Upgrade later for 90 days if hop evidence is the product they want. Do not pretend the body lives in MailerZ forever.
Agencies: paste the class in the subject. Leftover, HOLD, dest-5xx, empty-no-evidence. The next hire should not open the zone first.
After the customer resends and history is green, leftover check anyway. Helpful restores follow angry customers.
Envelope SRS on the stored copy is normal. Header From must still be the customer. If it is not, you are not debugging MailerZ.
Start free to have a hop to debug. A suite leftover with no MailerZ MX will never grow a hop row. Exclusive MX is the precondition, not the punchline.
Worked scenarios for missing customer email
A customer swears they emailed support@ at 3 p.m. Their screenshot shows a draft, not a sent item. Ask for the sent folder and a time in UTC. If they never got a 250 at their provider, MailerZ history will be empty for a reason. Do not republish MX to soothe the call.
History shows 250 toward Gmail and the owner cannot find it. Have them search all mail, then spam, then the exact subject. Destination folder is a class. It is not 'we never got it' until they searched. Self-send tests from the same Gmail still lie. Use a third mailbox for your own proof.
History is empty and public MX still lists Google next to MailerZ. The customer likely hit the leftover. Delete leftover MX. Do not recover a message that never arrived. Probe after exclusive. Then ask them to resend.
History shows HOLD. The address they used was a typo or an unpublished department name. Recover inside the store window if policy allows, or ask for a resend to the printed alias. Do not enable catch-all FORWARD as a permanent apology.
History shows 5xx with a reason. Read the reason. Unauthorized send-as is not inbound. Mailbox full at destination is not leftover MX. Classify before the second edit. RFC 5321 codes are cheaper than a DNS flap.
The customer used an old domain you no longer host. Their sent mail is real and your store is empty. That is a brand surface problem. Update the website. Do not invent a hop on a zone you do not publish.
Practice and anti-patterns in missing-mail tickets
Practice: time, exact To:, sent-folder proof, then history. Anti-pattern: apology plus MX republish in the first ten minutes.
Practice: one class per ticket—leftover, HOLD, 5xx, 4xx, destination folder, or no send. Anti-pattern: all classes at once so nobody can close it.
Practice: recover inside the window when you have the copy and policy allows. Anti-pattern: promising a copy after the clock. Fourteen days on Free. Ninety on paid. Quote /pricing.
Practice: third-mailbox probes for your own verification. Anti-pattern: the founder emails support@ from the destination Gmail and declares the system broken or fixed.
Practice: UTC. Anti-pattern: 'this afternoon' across three timezones and a store you search on the wrong day.
Practice: leave exclusive MX alone unless leftover is the class. Anti-pattern: adding the old host 'just in case' which creates the next missing-mail ticket.
Practice: do not paste customer bodies into Slack. Anti-pattern: a channel full of invoices that outlive the store and the need.
Operator closeout when the customer is still on the line
Closeout states the class in one sentence the customer can repeat. 'Your mail hit leftover Google MX' is a sentence. 'We're looking into DNS' is not.
If they must resend, say why: no hop, expired store, or destination rejected. Give the printed address. Do not give a new unpublished alias in the same breath as an apology.
If you recovered, say you placed it in the destination and they should search the unique subject. Do not claim inboxing. Placement is their provider.
If leftover MX was the class, closeout includes two-resolver listings after delete and a probe ID. The customer can resend after that proof, not before.
Agencies send the packet to the client owner, not only to the end customer. The client will otherwise flap MX again tonight.
Log UTC. Name the operator. If the same customer returns tomorrow, the next person should not start from zero.
Edge cases in 'they emailed us' stories
They emailed a person@ alias you remapped last week and their sent mail is real. History will show the new destination. The old human is searching the old mailbox. Teach the remap, do not chase MX.
They used a plus tag at a mailbox you do not control. You cannot find it in MailerZ. That string never hit your MX. Ask for the exact domain in the To: field.
A mobile client shows sent and the provider later bounced. Their UI lied for an hour. Ask for the bounce. Classify 4xx versus 5xx. Do not treat a bounce as a missing hop at MailerZ if MailerZ never saw it.
Two brands, one support phone number, two zones. They emailed the other brand. History on this zone is empty. Check the sister domain before you rebuild DNS.
A ticket system ate the forwarded copy with a rule. MailerZ 250 is real. The tool dropped it. Proof is hop plus the ticket filter log. Do not punish MX for a Zapier.
Legal wants every missing-mail ticket kept forever. Your hop store will not do that. Export the transcript they need inside the clock. Buy an archive if the business requires one.
Field notes from support floors
The phrase 'I emailed you' often means 'I typed something in a form' or 'I thought about emailing you.' Sent-folder proof is kindness, not distrust.
The phrase 'check your spam' is rude if you have not classified. Use it after a 250 toward their destination, not as the opening line.
Agencies that keep a launch-day probe ID close tickets faster. You compare shapes. Without that, every incident is archaeology.
HOLD surprises people who printed five addresses and created three. The two unpublished names are where 'customers' go to die. Print what you map. Map what you print.
Use /troubleshooting and /tools when the console and the customer disagree. A leftover MX checker is faster than a debate.
No inboxing percentage in the reply. You can prove a hop. You cannot promise a tab.
Handoff memo for the next missing-mail ticket
Printed aliases, HOLD policy, store window, leftover MX status, and where history lives in the product. If that memo is missing, the next agent will flap DNS.
Name the destination owners. If support@ lands in a shared Gmail nobody opens, the mail is not missing. It is unread. That is a staffing class.
Name who may recover from HOLD. That is break-glass adjacent. Do not let every agent open every body.
Link /email-forwarding and /docs. Inbound proof is not send-as. If they wanted a reply from the domain, that is a paid SMTP chapter.
Free versus paid store clocks belong in the memo. Agents will promise ninety days on a Free domain unless you write the number.
Acceptance criteria for the debug
You have a class, not a vibe. The customer heard the class. The ticket shows time, To:, and a history screenshot or a written empty-history reason.
If leftover was the class, two resolvers are exclusive and a probe succeeded before you asked for a resend.
If HOLD was the class, you either recovered under policy or asked for a resend to a mapped name. You did not enable panic FORWARD.
If destination folder was the class, they searched all mail. You did not republish MX.
No Slack dump of the customer body. No invented SLA. No SOC 2 story in the reply.
UTC is in the log. The next agent can continue.
Operations review of missing-mail handling
Count last month's tickets by class. If leftover MX wins, your launch checklist is broken. If HOLD wins, your printed map is broken. If folder wins, your customer language is broken. Fix the winning class, not all classes with MX.
Review agent macros. 'We reset DNS' should not exist. 'Send the sent-folder screenshot and the exact address' should.
Review store clocks against /pricing. Training that says 'we can always pull it' is a defect.
Review exclusive MX on the top five domains every week, not only after a complaint. Complaints are a trailing indicator.
Review who can open HOLD. Volume of glass should be low. If every missing-mail ticket becomes a body dump, you have no control.
Start free on a lab and run a fake customer drill. The first real donor should not be the first classification practice.
Quarterly review of evidence habits
Did we flap MX on a folder-class ticket? Write the miss in the postmortem template you already have.
Did we promise a copy after expiry? Retract the macro. Quote the window.
Did website saves republish Google MX? Morning re-query. Assign the website owner.
Did we keep customer bodies in chat? Delete. Remind. Repeat.
Are printed addresses still mapped? Crawl the site. HOLD should not be the unpublished footer.
Author: MailerZ editorial, Secuno LLC. Review when provider UIs, pricing, or scope change.
Closing notes on evidence before apology
Missing customer email troubleshooting starts with time, address, and hop history. Classify leftover MX, HOLD, codes, destination folders, or no send. Recover inside the clock when you have the copy. Do not republish MX to perform empathy.
MailerZ gives you hops and a short store. It does not give you telepathy or an inboxing SLA. Use the hops. Keep exclusive MX. Keep the map aligned with what you print.
Start free, run one drill, and answer the next customer with a class. Product path: /troubleshooting, /tools, /email-forwarding.
Intake form that prevents DNS panic
Fields: customer domain they sent from, exact To: string, UTC time, unique subject if they can resend, sent-folder screenshot, their provider (Gmail/Outlook/other).
Internal fields: two-resolver MX, hop row yes/no, HOLD yes/no, SMTP line if any, destination folder if found, leftover list, one change planned.
If To: is a plus tag you never created: HOLD or empty. Create or tell them the real local-part. Do not flap MX.
If To: is old registrar mail: leftover or wrong hop. Delete leftover. Ask resend.
If time is outside Free 14 days: say hop row may be gone. Search destination. Do not invent store.
If they only self-sent: ask another provider. Gmail-to-Gmail can hide MX.
If exclusive MX and HOLD: create alias. Resend.
If exclusive MX and dest 5xx: attach line. Work Gmail. No MX edit.
If leftovers: delete. Resend. Do not FORWARD to compensate.
If no screenshot and no resend: park the ticket. DNS is closed.
If two staff edit MX and SMTP: stop. Restore causality. One object.
If Header From rewritten on the copy they forwarded: not our hop. We do not rewrite Header From.
If Agency: one client in the subject. No blended threads.
If sales wants a placement promise in the customer reply: remove it. Attach evidence.
After green resend: leftover check. Angry weeks spawn aspmx restores.
Caps and plans do not debug inbound. 550 send-as is another ticket.
Start free so there is a hop row to attach. Suite leftover without MailerZ MX will never grow that row.
Close with class in the first line. The next hire should not open Route 53 first.
FAQ
- What is the safest way to handle missing customer email troubleshooting?
- Ask for the time (UTC), the exact To: address, and a screenshot of their sent folder. Open MailerZ history. Classify leftover MX, HOLD, 5xx, 250-plus-folder, or no evidence. Recover inside the window if you have the copy. Do not republish MX first.
- Does this require a new mailbox?
- No. MailerZ is not IMAP. Keep Gmail or Outlook unless you need a suite for other reasons.
- Will it work with Gmail or Outlook?
- Yes as destinations. Self-send is not proof. Use a third mailbox and open original.
- What DNS records are involved?
- Exclusive MX, verification TXT, one SPF if you send-as. Leftover MX is a hard stop. Dashboard values only for sending.
- What should I test before production?
- A uniquely titled probe from an unrelated provider to each public alias. Confirm Header From and hop history.
Key takeaways
- Evidence before DNS.
- Classify leftover, HOLD, reject, folder, or no send.
- Store clock.
- Sent-folder ask.
- Do not recover old-host mail.
- Watch junk search.
- One class sentence.
- No catch-all make-good.
Conclusion
Believe the feeling. Prove the hop. Name the class. That is how “they emailed us” becomes an answer.
Start free and keep history inside the window so the next ticket has artifacts.