14-day vs 90-day vs 180-day message recovery: how much retention do you need?

Quick answer for email forwarding retention

Use fourteen days when the job is a first domain, an inbound probe, and a short retry loop. That is MailerZ Free: one domain, three aliases, one seat, a 14-day store, send-as disabled, SMTP and API disabled, and unrouted mail held or rejected only. Use ninety days when support tickets, vendor disputes, or destination outages regularly show up after the second week. Paid plans keep a 90-day store and add send-as within published hourly and monthly limits.

Email forwarding retention setup is therefore a plan choice plus a habit. The habit is looking at delivery history while the row still exists. The plan choice is Free versus Solo, Starter, Business, or Agency. Solo is $40 per year only. Starter is $8 monthly or $80 yearly. Business is $19 or $190. Agency is $39 or $390. Confirm current numbers on MailerZ pricing. Those figures are capacity and store length, not a deliverability contract.

Internet mail does not require the receiver to keep a copy for you. IETF RFC 5321 — Simple Mail Transfer Protocol is a transport standard. Once the destination SMTP server says yes, that server owns the next fate. MailerZ can record the response and keep a copy for the published window. Gmail can still hide the message. Outlook can still junk it. A 90-day row that says “accepted” and an empty inbox can both be true.

MailerZ is a delivery layer operated by Secuno LLC. Envelope MAIL FROM can use Sender Rewriting Scheme. Header From, Subject, Date, Message-ID, body, and MIME are never rewritten. It is not Google Workspace, not IMAP, and not an open relay. Unhosted or unauthorized recipients get SMTP 550 / 550 5.7.1. The recovery surface in product language is delivery recovery.

Email forwarding retention best practice is boring: keep the mailbox you search in a year, keep leftover MX deleted, probe from a different mailbox, and pick a window that matches how late your tickets arrive. If you need years of search, you need the destination product’s archive features, not a longer MailerZ slider that does not exist.

The user problem and the decision criteria

The usual ticket arrives on day eighteen. A customer swears they sent a contract. The founder searches Gmail. Nothing. The forwarder they used for free already dropped the body. There is no SMTP reason, only a shrug. They upgrade something, anything, and still cannot say whether the destination refused the hop last month.

The other ticket arrives on day three. The founder wants “unlimited retention” because a competitor slide said enterprise. They do not have a legal hold process. They do not search MailerZ. They search Gmail. They are buying a window they will not open, while leftover MX still splits inbound.

People confuse three stores. The destination mailbox is where humans read. The MailerZ window is where hops and recoverable copies live for a published number of days. The sending server’s own logs are a third place you do not control. Email forwarding retention guide posts that pretend one slider replaces all three are selling a story.

Unknown recipients add a fourth pile. Free holds unknown mail instead of silently forwarding guesses. Paid plans can forward catch-all when you choose that exposure. Held mail still ages out with the window. If you never look at the hold queue, a longer window only stores unread noise.

Weekend operators feel the 14-day edge first. A message accepted Friday night, noticed the following Monday, is still inside Free. The same message noticed after a two-week vacation is gone. That is not a product defect. It is the published window. If the person who reads history takes holidays longer than the store, you need 90 days or a deputy, not a angry ticket to support.

Shared inboxes make the lag worse. Ten people assume someone else opened the hop. Nobody did. The row expires. A 90-day window does not fix the RACI. It only keeps the evidence available while you argue. Write who looks at delivery history before you pay for the longer pack.

Technical mail flow

A sender looks up MX and delivers. MailerZ accepts for a verified domain and a configured recipient, or holds unknown mail on Free. It then attempts the destination. The destination returns an SMTP reply. That reply is the fact. The window is how long you can still see the fact and, when the product allows, recover a copy.

What the store is

The store is a time-bounded copy tied to delivery evidence. It exists so you can see why a hop failed and, inside the window, recover. Opening a body for break-glass recovery can create an audit row. That is a control, not a marketing badge. Do not treat it as a second IMAP login. There is no MailerZ webmail reading pane that replaces Gmail.

What the store is not

It is not a substitute for Gmail’s archive, Vault, or Outlook litigation hold. MailerZ does not claim SOC 2, ISO 27001, or HIPAA. It does not publish an uptime SLA. A 90-day window is longer than 14. It is not infinite. It is not searchable the way a mailbox is searchable a year later.

Authentication and retention

SPF, DKIM, and DMARC evaluate the sending identity. IETF RFC 7208 — Sender Policy Framework (SPF), IETF RFC 6376 — DomainKeys Identified Mail (DKIM), and IETF RFC 7489 — Domain-based Message Authentication, Reporting, and Conformance (DMARC) are the specs. They do not keep your mail. A DMARC reject at a receiver can appear in history as a destination refusal. Keeping that row for 90 days helps the postmortem. It does not change the policy that caused the bounce.

Header-preserving forward matters here too. If a different tool rewrote From, you will debug the wrong identity. MailerZ does not rewrite Header From. Envelope SRS is the rewrite that is allowed. Features are listed on MailerZ features. None of those bullets are an inbox guarantee.

Retries and the clock

A destination 4xx can mean “try later.” A 5xx usually means stop. The store keeps the message so you can decide whether a retry is honest. Retrying a 550 5.7.1 unauthorized From is not persistence. It is repeating a policy failure. Retrying a mailbox-full 452 might be honest if the destination recovered. The window is how long that choice remains available. Fourteen days covers a short outage. Ninety days covers a destination that was down while your team was also down.

Self-send tests waste the clock. Gmail can accept a message to itself without showing the forward path. You then have a history row that looks quiet, or no row, and a false sense that retention is empty. Probe from another provider. That is in the docs for a reason. The window cannot store a hop that never left the destination’s local shortcut.

Time zones surprise people on day fourteen. A message stored at 23:50 UTC on day one may fall off while a local calendar still says “two weeks.” Do not design a customer promise on the last hour of Free. If the relationship needs a predictable month of evidence, you are describing paid 90-day store, not a reminder bot on day thirteen.

Step-by-step setup and decision path

Pick the window after you name the ticket lag. Do not pick it from a fear of deletion in the abstract.

1. Write how late a real ticket arrives

   If you always notice missing mail the same afternoon, fourteen days is generous. If finance asks about an invoice from last month, fourteen days is already gone. Use last quarter’s tickets, not a guess.

2. Separate archive from recovery

   If counsel wants seven years, stop this article and buy a mailbox archive. If you want to replay a 550 and show a timestamp, stay.

3. Add one domain and prove inbound

   Verification TXT, one MX set, leftover MX deleted, uniquely titled probe from another provider. Confirm Header From and a history row. Self-send lies. DNS help lives on troubleshooting and DNS diagnostics.

4. Stay on Free while the path is the question

   The 14-day store is the right size for a cutover week. Watch the hold queue. Do not print every role address yet.

5. Upgrade when the window or send-as is the question

   Paid 90-day store arrives with the paid plans. Send-as is the other reason to pay. You can need one, the other, or both. Copy SMTP values only after inbound works. Gmail’s labels are in Google Gmail Help — Send mail from a different address.

6. Practice a recovery while a test message is still inside the window

   Do not wait for a production emergency to learn where the button is. Sanitize screenshots. Do not mail SMTP passwords to anyone.

7. Write down when rows age out

   Fourteen days is two weeks. Ninety days is a quarter. Calendar the expectation so nobody asks for a March hop in July.

Field maps live in the MailerZ documentation. The docs will not invent a year-long store. If a page ever disagrees with pricing, pricing wins for limits.

Failure modes and proof

Retention failures are usually category errors. People look in the wrong store, or they look after the window. Proof is a history row with a timestamp and a remote SMTP response, plus the destination inbox if the hop succeeded.

Do not publish message bodies from recovery into a ticket system without a retention rule of your own. The MailerZ window ending does not erase copies you pasted elsewhere. That is your process.

MailerZ workflow and product boundary

MailerZ is a custom-domain email delivery layer operated by Secuno LLC. Point MX at MailerZ. Mail for a hosted, verified domain can land in Gmail or Outlook. Paid plans add authenticated SMTP so the same domain can send and reply. The live site is mailerz.net. The app is mail.mailerz.net.

Positioning for this article: show exact SMTP outcomes and recover failures during the retention window. The product appears after you admit the mailbox is elsewhere. MailerZ does not become a records suite because the paid window is longer.

What MailerZ does in this workflow

• Accept inbound mail for verified domains and configured recipients.
• Preserve Header From on the forward.
• Hold or forward unknown recipients according to plan and settings.
• Store messages for 14 days on Free or 90 days on paid plans for recovery and evidence.
• Record delivery history with destination responses.
• Send through authenticated SMTP from approved identities on paid plans.
• Surface leftover MX as a diagnostic, not as a health score.

What MailerZ does not do

• Replace Gmail or Outlook with IMAP or webmail.
• Rewrite header From, Subject, Date, Message-ID, body, or MIME.
• Offer send-as on Free.
• Promise inbox placement, uptime SLAs, or review counts.
• Claim SOC 2, ISO 27001, or HIPAA. Wording lives on Security and Trust Center.
• Act as an open relay. Unauthorized send gets 550 / 550 5.7.1.
• Provide multi-year legal hold or eDiscovery.

Outbound hops on paid plans also produce history. Monthly outgoing caps still apply. A 90-day store full of campaign attempts is a sign you used the wrong product. Use a campaign platform for lists.

Cost, alternatives, and trade-offs

The retention upgrade is bundled with paid plans. You do not buy “90 days” as a standalone add-on in the published cards. You buy Solo or a higher plan and receive the 90-day store plus the other paid capacities. If you only wanted send-as, you still get 90 days. If you only wanted 90 days, you still get send-as capacity you may not use. That is the pack.

Registrar forwarding toggles often keep nothing useful. Cloudflare Email Routing is inbound routing; read their docs for what they store. ImprovMX is the closest commercial forwarding class; compare live plan retention instead of memory. Google Workspace retention is a suite feature on a hosted mailbox. Different object.

Operational cost is the habit. A 90-day store you never query is a more expensive 14-day store. Calendar a weekly hold-queue glance during the first month. Then decide if the paid window is earning its keep.

If leftover MX is still published, neither window matters. Half the senders never reach you. Fix MX first. Retention second. Send-as third.

Compare store length to destination retention only after you admit they are different clocks. Gmail keeps mail for as long as that account exists, subject to Google’s own policies. MailerZ Free deletes its copy after fourteen days. Paying for Workspace so that Gmail “keeps more” does not lengthen the MailerZ window. Paying for MailerZ Business does not lengthen Gmail. Buy the clock you are actually reading.

Agencies with many domains should not assume one window fits every client. A client who disputes invoices monthly needs 90 days on that domain’s paid plan. A parked brand that only receives a weekly probe can live on Free until send-as is required. Agency’s domain ceiling is a capacity number. It is not a retention policy per hostname. You still choose the plan that owns each domain.

Outbound recovery is easy to over-request. If a paid send fails with 550 5.7.1, the history row is the artifact. You do not need ninety days to read a rejection you caused this afternoon. You need ninety days when the recipient claims last month’s outreach never happened and you must show the destination response. Match the window to that claim, not to a fear of all possible claims.

FAQ

Key takeaways

• Email forwarding retention is a recovery window. The mailbox is still Gmail or Outlook.
• Fourteen days fits cutover proof and fast retries. Free includes that window.
• Ninety days fits late tickets and quoted SMTP reasons. Paid plans include that window.
• Neither window is legal hold, IMAP, or an inbox SLA.
• History can say delivered while the tab says empty. Both can be true.
• Leftover MX drops mail before retention can see it.
• Held unknown mail ages out too. Look at the queue.
• Send-as and store length travel together on paid cards. Confirm /pricing.
• Header From stays. Envelope SRS is the allowed rewrite.
• MailerZ is not SOC 2 and not a bulk sender.

Conclusion and next action

If you came here for 14-day vs 90-day vs 180-day message recovery, measure ticket lag, not fear. Prove inbound on the short window. Pay for the long window when you still need the hop after two weeks, or when send-as is already required. Keep the archive in the inbox product you actually search.

MailerZ fits when you want that evidence attached to forwarding and paid SMTP. It does not fit when you need a hosted mailbox for every user, a certified compliance report, or campaign-scale sending. Start on Free if you only need to watch inbound. Move to Solo or another paid plan when the store, aliases, or From identity require it.

Next action: add one domain, create one alias, and send a uniquely titled message from a mailbox that is not the destination. Open delivery history the same day. Decide whether fourteen days would have covered last quarter’s worst ticket. If the worst ticket was thirty days late, you already know the plan shape. The register path is one domain, not an archive migration. The MailerZ documentation has the field maps if a label in the desk app is unclear.

Review quarterly, or sooner if MailerZ store length, plan limits, or DNS guidance changes. Author: MailerZ editorial, Secuno LLC.