What does "553 SPF (Sender Policy Framework) domain authentication fail" mean?
The "553 SPF (Sender Policy Framework) domain authentication fail" bounce means the recipient's mail gateway, Broadcom's Symantec Email Security.cloud, rejected your message because it failed an SPF check against a domain that publishes a hard fail (-all) policy. The sending IP is not authorised in the SPF record of your envelope sender domain. The fix is almost always on the sender's side: correct the SPF record.
The text is documented by Broadcom in its knowledge base article Email delivery failure due to SPF Domain Authentication Fail (article 175407), where the error appears as 553-SPF (Sender Policy Framework) domain authentication fail. Many people search for it together with "G Suite" or "Google" because the bounce lands in a Gmail or Google Workspace inbox, but Google did not generate it. Google's own Gmail SMTP errors and codes reference has no such entry; its 553 codes are about invalid addresses. The rejection happened at the recipient's gateway, and Gmail simply delivered the non-delivery report (NDR) back to you. For SPF fundamentals, see the SPF guide.
Key takeaways
- The bounce is issued by Symantec Email Security.cloud when its Spoofed Sender Detection "Use SPF" setting is set to block and the sender's SPF result is a hard fail.
- Broadcom states that a softfail (
~all) policy does not trigger this rejection. Only-allcombined with an SPF failure does. - The check runs on the envelope sender (MAIL FROM, shown as Return-Path), not the visible From address, and aligned DKIM does not rescue the message.
- The usual causes are a sending platform missing from the SPF record, mail forwarded through an intermediate server, or a broken record that never passes.
- The sender's DNS administrator fixes it; the recipient can only add a temporary exception.
Which system sends this bounce?
Symantec Email Security.cloud, a cloud email gateway sold by Broadcom, sends it on behalf of the recipient's organisation. If your client's mail bounces with this text, the organisation you were writing to routes its inbound mail through that service, and it judged your message as spoofed based on your SPF record.
Broadcom's article describes the mechanism. The Spoofed Sender Detection setting "Use SPF", when configured to Block and Delete, checks inbound mail against the SPF policy of the envelope sender. If the sending domain publishes a hard fail policy and the message fails, the message is blocked and the sender receives a 5xx error. The same article lists Email Security.cloud as the environment.
Broadcom documents a sibling text as well. Its article Outbound Emails Rejected Due to DMARC Failure shows a Microsoft 365 NDR reading Remote Server returned '553-DMARC domain authentication fail', for messages that fail SPF and DMARC when the sender's domain publishes p=reject. The SPF variant is the one this guide covers; the DMARC variant is closer to the problems in our 554 5.7.5 DMARC error guide.
What do the 553 code and the SPF result mean?
The 553 reply code is a permanent rejection, and the SPF result behind it is "fail". RFC 5321, Section 4.2.2 defines 553 as "Requested action not taken: mailbox name not allowed (e.g., mailbox syntax incorrect)". Here the gateway uses it for a policy rejection of the sender address rather than a syntax problem.
Two details are worth knowing when you read the code:
- The first digit is 5. Under RFC 5321, Section 4.2.1, a 5yz reply means the command was rejected and the action did not happen. Your server will not retry. Resending the same message unchanged gives the same result.
- 553 is not what the SPF standard suggests. RFC 7208, Section 8.4 says a server rejecting on an SPF fail "SHOULD use an SMTP reply code of 550". Gateways are free to choose differently, which is why searching only for "550 SPF" misses this bounce.
SPF itself can return seven results, defined in RFC 7208, Section 2.6. Only some of them lead to this rejection:
| SPF result | Record ending or cause | RFC 7208 meaning | Triggers this 553? |
|---|---|---|---|
pass |
IP matches a mechanism | Client is authorised | No |
fail |
IP falls through to -all |
"An explicit statement that the client is not authorized" | Yes |
softfail |
IP falls through to ~all |
Probably not authorised; receivers "SHOULD NOT reject the message based solely on this result" (Section 8.5) | No, per Broadcom |
neutral |
?all, or no all |
Domain makes no assertion | No |
none |
No SPF record found | Nothing to check | No |
temperror |
DNS timeout | Transient error | Not documented |
permerror |
Broken record | Record cannot be interpreted | Not documented by Broadcom; other receivers may reject |
The practical consequence: a domain ending in -all is exposed to this gateway every time any legitimate source is missing from its record. SPF softfail vs hardfail with DMARC covers that trade-off in full.
Why does a message fail SPF in the first place?
SPF fails when the IP address that connected to the recipient's gateway is not listed, directly or through an include:, in the SPF record of the envelope sender domain. The record is correct for the senders someone remembered, and wrong for the one that just bounced.
The common causes, roughly in the order an MSP meets them:
- A sending platform was never added. A new CRM, ticketing system, invoicing tool or email service provider (ESP) sends with the client's domain in MAIL FROM, but nobody added its
include:. - An on-premises server or device sends directly. A scanner, line-of-business application or old Exchange relay sends from the office IP, which is not in the record.
- The mail was forwarded. A mailbox or alias forwards to a recipient behind the gateway. The forwarder's IP is not in your record, so SPF fails at the second hop. RFC 7208, Section 10.2 acknowledges SPF can "return 'fail' for mail that is ultimately legitimate" in exactly this case.
- A migration changed the outbound path. Mail now leaves through a new tenant, a new outbound gateway or a different region, and the old record still describes the previous route.
- The record is broken. More than 10 DNS-querying terms, two records at the same name, or a typo produce
permerror, so nothing ever passes.
The last cause deserves its own check. RFC 7208, Section 4.6.4, says implementations "MUST limit the total number of those terms to 10", and Section 4.5 says that if a domain has more than one SPF record, check_host() produces permerror. Broadcom's article only documents hard fail as the trigger, but a broken record leaves you with no SPF pass anywhere, which is a deliverability problem at every receiver. See SPF too many DNS lookups and multiple SPF records.
How do I read the bounce?
Find three things in the NDR: the domain that was checked, the IP that connected, and the server that rejected. The 553 text itself gives none of them, so look at the diagnostic details and the original message headers attached to the bounce.
An NDR from a Google Workspace or Microsoft 365 mailbox will look roughly like this (illustrative, example values):
Remote server returned '553-SPF (Sender Policy Framework) domain authentication fail'
Reporting-MTA: dns; mail.example.com
Final-Recipient: rfc822; accounts@example.co.uk
Action: failed
Status: 5.0.0
Remote-MTA: dns; mx1.gateway.example.net
Read it like this:
- Remote-MTA or the rejecting host name tells you which gateway refused the message. That host belongs to the recipient's mail routing, so check their MX record (below) if the name is unclear.
- The envelope sender is the Return-Path of your original message. For mail sent from a user's mailbox it is normally the user's own domain; for an ESP it may be a bounce subdomain such as
bounce.example.com. - The connecting IP is the last hop before the gateway: your provider's outbound server, or a forwarder if the mail was relayed.
Confirm the recipient's MX to see which gateway they use:
dig +short MX example.co.uk
nslookup -type=mx example.co.uk
Then open the headers of a copy of the same message that reached a mailbox you control, and read the Authentication-Results header. A message from a forgotten platform looks like this (illustrative):
Authentication-Results: mx.example.net;
spf=fail (sender IP is 198.51.100.40) smtp.mailfrom=example.com;
dkim=pass header.d=example.com;
dmarc=pass header.from=example.com
This is the case that surprises people: DKIM passed and aligned, DMARC passed, yet a gateway that rejects on SPF fail alone would still bounce it. RFC 9989, Section 7.1 warns about this directly: a hard fail on -all "could cause a message to be rejected early in the SMTP transaction, before any DMARC processing takes place". Our email header analyzer puts the three results side by side, and how to read email headers explains every field.
How do I fix a 553 SPF domain authentication fail?
Add the missing sending source to the SPF record of the envelope sender domain, keep the record valid, and resend. Broadcom recommends exactly this as the primary fix: the sender's DNS administrator corrects the SPF record so it includes the sending IP address.
Step 1: Look up the current record
dig +short TXT example.com | grep spf1
nslookup -type=txt example.com
Resolve-DnsName -Name example.com -Type TXT | Where-Object { $_.Strings -match 'v=spf1' }
You should see exactly one record starting with v=spf1. If you see two, merge them before doing anything else.
Step 2: Identify the source that failed
Match the connecting IP from the bounce against the client's known platforms. A reverse lookup often names the provider:
dig +short -x 198.51.100.40
If the IP belongs to a forwarder rather than to your client's platform, skip to the forwarding section below: adding a third party's forwarding server to your record is not the fix.
Step 3: Add the platform's documented mechanism
Use the vendor's published include: rather than individual IPs where one exists. For Google Workspace that is include:_spf.google.com; for Microsoft 365 it is include:spf.protection.outlook.com. Our guides to the Google Workspace SPF record and the Microsoft 365 SPF record show both in context.
Before:
example.com. TXT "v=spf1 include:spf.protection.outlook.com -all"
After (office relay and an ESP added):
example.com. TXT "v=spf1 include:spf.protection.outlook.com include:spf.esp.example ip4:203.0.113.25 -all"
Step 4: Count the lookups
Each include, a, mx, ptr, exists and redirect counts towards the limit of 10, including those nested inside included records. ip4 and ip6 do not count. If the new include pushes you over 10, remove dead entries or move a platform to a subdomain Return-Path before considering flattening, as explained in SPF flattening for MSPs. Check the live result with our SPF analyzer.
Step 5: Decide whether to keep -all
Broadcom offers switching the record to softfail as an interim measure while the record is corrected, and states that a softfail policy does not trigger this rejection. That is a legitimate choice if every source signs with aligned DKIM and the domain is at DMARC enforcement, because DMARC treats softfail and fail the same way. It is not a substitute for adding the missing source.
Step 6: Resend and confirm
After the TTL expires, resend the original message. A 5xx rejection is final, so the bounced copy will never be delivered by retry.
What can the recipient do while the sender fixes it?
The recipient's administrator can add a temporary exception, but Broadcom treats it as a workaround, not a fix. In the Email Security.cloud portal (ClientNet), Broadcom lists these options for time-sensitive mail:
| Option | Effect | Broadcom's position |
|---|---|---|
| Add the sender or domain to Anti-Spam Approved Senders with "Spoof" selected | Bypasses only the SPF and DMARC checks for that sender | Temporary measure |
| Change "Use SPF" from Block and Delete to Quarantine or Tag a header | Message is held or tagged instead of rejected | Temporary measure |
| Disable "Use SPF" | No SPF-based blocking at all | Not recommended |
Broadcom notes that portal changes can take up to one hour to propagate, after which the sender must resend. If your MSP manages the recipient's gateway rather than the sender's DNS, this is your lever; push the sender to correct their record at the same time.
How does forwarding cause this bounce, and what is the fix?
Forwarding causes it when a message passes through an intermediate server that keeps your domain in MAIL FROM. The gateway sees the forwarder's IP, your record does not list it, and -all turns that into a hard fail. You cannot fix that by adding every forwarder on the internet to your record.
The realistic options are:
- Ask the forwarder to rewrite the envelope sender (Sender Rewriting Scheme), so SPF is evaluated against the forwarder's own domain.
- Have the recipient use direct delivery instead of an alias or auto-forward that hops through another provider.
- Use
~allon the sending domain, relying on aligned DKIM for DMARC, so gateways that follow RFC 7208, Section 8.5 do not reject on SPF alone.
DMARC and email forwarding explains which forwarding setups break SPF, which break DKIM, and where ARC helps.
How does DMARC relate to this SPF bounce?
DMARC is not involved in the SPF variant of this error: the gateway rejects on the raw SPF result before DMARC matters. But your DMARC setup decides how safe it is to change the SPF qualifier, and DMARC aggregate reports are the best way to find the missing source before it bounces.
DMARC passes when SPF or DKIM produces a pass for a domain aligned with the From address, as defined in RFC 9989. DKIM alignment is the safer path for three reasons:
- DKIM survives forwarding, provided the message is not modified, while SPF breaks at the first relay.
- DKIM is tied to your domain, through the
d=tag, wherever the platform sends from, so it does not depend on an IP list you have to maintain. - SPF has a hard ceiling of 10 lookups, and every new platform eats into it.
Once every legitimate source signs with aligned DKIM and the domain is at p=quarantine or p=reject, ~all loses little protection and removes the exposure to SPF-only rejections like this one. There is a reporting catch, though. RFC 9989, Section 7.1, notes that "messages rejected early in the SMTP transaction will never appear in aggregate DMARC reports". A gateway that rejects on SPF at the envelope stage may never report the message, so the NDR and the user's helpdesk ticket can be your only evidence. DMARC alignment explained and why DMARC fails when SPF passes cover the alignment side.
How should an MSP handle this across a client portfolio?
Treat each 553 bounce as evidence of an unlisted sender or a forwarding path, and fix the class of problem, not just the one record. The same gap usually affects every receiver that enforces SPF, not only the gateway that happened to reject.
- Log the full NDR text in the ticket, including the rejecting host and any IP. "Email bounced" is not enough to act on.
- Identify which side you manage. If your client is the sender, fix DNS. If your client is the recipient on Email Security.cloud, consider a temporary approved-sender exception and tell the sender.
- Check the domain's aggregate reports for other sources with SPF
failorsoftfail. DMARC aggregate reports explained shows where to look. - Audit every client domain ending in
-allfor lookup count, duplicate records and missing platforms. - Make aligned DKIM the standard on every platform that supports it, so a future SPF gap does not break DMARC.
- Change DNS under change control with client approval, then send a test and confirm
spf=passin the headers.
A multi-tenant platform such as DMARC AI for MSPs processes aggregate reports per client domain and classifies sending sources, which turns step 3 into a per-client view. For UK clients, keeping SPF, DKIM and DMARC accurate is part of the NCSC's email security and anti-spoofing guidance, and a source missing from SPF is the kind of gap that guidance asks you to close.
Which related bounce codes should I know?
Other receivers report SPF and alignment failures with their own codes. If you are triaging several bounces for the same domain, these are the ones most likely to appear alongside this one:
| Bounce | Issued by | What failed | Guide |
|---|---|---|---|
553-SPF (Sender Policy Framework) domain authentication fail |
Symantec Email Security.cloud | SPF hard fail on envelope sender | This guide |
553-DMARC domain authentication fail |
Symantec Email Security.cloud | SPF and DMARC fail under p=reject |
Why DMARC fails when SPF passes |
550 5.7.26, 550 5.7.27 |
Gmail | No authentication, SPF hard fail or DMARC policy | Gmail 550 5.7.26 bounce codes |
550 5.7.515 |
Outlook.com | High-volume sender authentication requirements | Microsoft 550 5.7.515 |
554 5.7.5 |
Various receivers | DMARC record could not be evaluated | 554 5.7.5 permanent error evaluating DMARC policy |
SPF permerror |
Any SPF verifier | Too many lookups or duplicate records | SPF too many DNS lookups, multiple SPF records |
If the client sits behind its own gateway as well, how secure email gateways affect DMARC, SPF and DKIM explains how an extra hop changes outbound SPF.
FAQ
Is the 553 SPF domain authentication fail error a Google or G Suite error?
No. Broadcom documents it as a Symantec Email Security.cloud rejection. It often arrives in a Gmail or Google Workspace inbox because that is the sender's mailbox, which returns the NDR. Google's Gmail SMTP error reference contains no such message; Gmail reports SPF failures with codes such as 550 5.7.26 and 550 5.7.27.
Will DKIM stop this bounce?
Not on its own. Broadcom describes the check as an SPF check against the envelope sender, rejecting on hard fail. A valid, aligned DKIM signature helps DMARC and other receivers, but this gateway setting looks at SPF. Fix the SPF record, or move to ~all once aligned DKIM covers every source.
Who has to fix it, the sender or the recipient?
The sender. The rejection reflects the sender's own SPF record, so the sender's DNS administrator must add the missing source. The recipient can add a temporary approved-sender exception in the gateway portal, which Broadcom positions as a stopgap for time-sensitive mail, not a permanent fix.
Does changing -all to ~all fix it immediately?
Broadcom states that a softfail policy does not trigger this rejection, so the bounce should stop once receivers see the new record, subject to its TTL. It hides the real problem, though. Add the missing source as well, and only keep ~all long term if aligned DKIM and DMARC enforcement are in place.
Why don't I see these failures in my DMARC reports?
Because the gateway can reject before the message data is transferred. RFC 9989, Section 7.1, notes that messages rejected early in the SMTP transaction never appear in aggregate reports. Use the NDRs and helpdesk tickets as your evidence, and use aggregate reports to find the same source failing SPF elsewhere.
Final thoughts
The "553 SPF (Sender Policy Framework) domain authentication fail" bounce is not a mystery error from Google. It is a Symantec Email Security.cloud gateway enforcing the sender's own -all policy against a source the record does not list. Find the connecting IP, add the platform's documented mechanism to the envelope sender's record, keep it under 10 lookups, and resend. Then make aligned DKIM the standard across the portfolio, so the next missing SPF entry costs a warning in a report rather than a bounced invoice.