How do you set up DMARC for Brevo?
Brevo DMARC setup works through domain authentication. You publish a brevo-code TXT record to prove ownership, two DKIM CNAME records (brevo1._domainkey and brevo2._domainkey) or one older DKIM TXT record, and a single DMARC TXT record at _dmarc. Brevo then signs mail with your domain, DKIM aligns with the From address, and DMARC passes.
Brevo, formerly Sendinblue, is a European email service provider (ESP) that MSPs (managed service providers) meet often at UK and Dutch SMB clients. Marketing teams sign up on their own, and the MSP only finds out when Brevo appears as a failing source in DMARC (Domain-based Message Authentication, Reporting and Conformance) reports. This guide lists exactly which records Brevo asks for, as documented in the Brevo Help Center as of November 2026, and explains what each record means for alignment. For protocol background, start with the DMARC guide.
Key takeaways
- Brevo asks for three things to authenticate a domain: a
brevo-codeTXT record, DKIM (two CNAMEs or one TXT) and a DMARC TXT record. SPF and MX are not required for authentication. - On shared IPs, the return path is Brevo's own
brevosend.com, so SPF passes for Brevo and does not align. DMARC depends on aligned DKIM unless you add a branded subdomain. - Brevo's suggested DMARC record is
v=DMARC1; p=none; rua=mailto:rua@dmarc.brevo.com. Merge its reporting address into the client's existing record instead of adding a second one. - Unauthenticated or free-mail senders get their From address rewritten to a Brevo domain such as
brevosend.com. - Domains already at
p=quarantineorp=rejectmust be authenticated before Brevo lets you create a sender on them.
Which DNS records does Brevo ask for?
Brevo asks for a Brevo code TXT record, DKIM records and a DMARC TXT record. Its domain authentication FAQ lists DKIM as "1 TXT or 2 CNAME". Brevo's main article states that SPF and MX records are not required to authenticate a domain and are only provided when you set up a dedicated IP.
| Record | Type | Host | Purpose |
|---|---|---|---|
| Brevo code | TXT | Root of the sending domain (@) |
Proves you control the domain |
| DKIM 1 | CNAME | brevo1._domainkey |
Points to a Brevo-hosted public key (2048-bit by default) |
| DKIM 2 | CNAME | brevo2._domainkey |
Second key, which lets Brevo rotate keys |
| DKIM (older accounts) | TXT | Selector shown in Brevo, often mail._domainkey |
Static key, 1024-bit by default |
| DMARC | TXT | _dmarc |
Policy and reporting address |
The host names brevo1._domainkey and brevo2._domainkey come from Brevo's domain authentication troubleshooting page. The values are unique per account, so always copy them from Settings > Senders, Domains, IPs > Domains. Brevo's own domain shows the pattern in public DNS: its CNAMEs point to targets of the form b1.<domain-with-dashes>.dkim.brevo.com.
A typical zone for a client sending from example.co.uk looks like this (values illustrative):
example.co.uk. TXT "brevo-code:0123456789abcdef0123456789abcdef"
brevo1._domainkey.example.co.uk. CNAME b1.example-co-uk.dkim.brevo.com.
brevo2._domainkey.example.co.uk. CNAME b2.example-co-uk.dkim.brevo.com.
_dmarc.example.co.uk. TXT "v=DMARC1; p=none; rua=mailto:dmarc@example.co.uk,mailto:rua@dmarc.brevo.com"
Should you use the CNAME or the TXT DKIM option?
Use the two CNAME records whenever Brevo offers them. Brevo states that CNAME records use a 2048-bit key by default, while TXT records default to 1024-bit. With CNAMEs, Brevo hosts and rotates the keys, so the MSP never touches the zone for a key change.
To move an older account from TXT to CNAME, Brevo says to add the two CNAMEs, keep the existing TXT record unchanged until the domain shows as authenticated, then click Authenticate this email domain. Only then remove the old TXT record. The DKIM selector guide explains why two selectors make rotation safe.
Does Brevo still need an SPF include?
Not for shared IP sending. Brevo's documentation says SPF is not required to authenticate a domain, and its sender requirements page lists SPF as handled during dedicated IP setup. Brevo's SPF merge article still shows include:spf.brevo.com in an example, so you will find it in many client records.
On shared IPs, that include changes nothing for DMARC, because SPF is checked against Brevo's return path, not the client's domain. It only spends one of the ten DNS lookups SPF allows under RFC 7208. Check with the client before removing it, in case a dedicated IP setup relies on it, and see SPF too many DNS lookups for budgeting.
Why does Brevo pass DMARC through DKIM and not SPF?
Because SPF authenticates the envelope sender, and Brevo uses its own domain there by default. Brevo's branded subdomain FAQ states that, without a branded subdomain, the mailed-by and return-path domain is brevosend.com. SPF passes for brevosend.com, which publishes v=spf1 include:spf.sendinblue.com -all, but it never matches the client's From domain.
RFC 9989, the DMARC standard that obsoletes RFC 7489, says a message passes when SPF or DKIM passes and aligns with the From domain. One aligned mechanism is enough. Once the brevo1 and brevo2 records are live, Brevo signs with d=example.co.uk, and DKIM carries DMARC on its own.
In aggregate reports, authenticated Brevo mail on shared IPs typically looks like this:
| Field | Value | Aligned? |
|---|---|---|
| SPF domain | brevosend.com |
No |
| DKIM domain | example.co.uk |
Yes |
| DMARC result | pass |
That is a healthy result. The details of relaxed and strict matching are in DMARC alignment explained.
What does a Brevo branded subdomain change?
A branded subdomain moves the return path, tracking links and image links onto the client's domain. Brevo's domain setup guide says this gives full SPF alignment, so DMARC has two independent signals instead of one.
In Brevo's newer setup flow, the branded subdomain is a CNAME such as mail.example.co.uk that links to Brevo for SPF and the return path, plus CNAMEs for image and tracking links. Brevo describes up to seven records in total, or an NS delegation option where Brevo manages the subdomain records. Brevo says this flow is rolling out gradually, so check what the client's account shows. With NS delegation, record who owns that delegated zone in your DNS documentation: changes there no longer pass through the MSP.
Does Brevo require a DMARC record?
Brevo asks for a DMARC TXT record as part of domain authentication, and strongly ties it to mailbox provider rules. Its DMARC FAQ says Brevo only provides values for a p=none policy and recommends v=DMARC1; p=none; rua=mailto:rua@dmarc.brevo.com. Brevo also warns that mail to Outlook.com, Hotmail and Live addresses may go to spam or be rejected without DMARC.
Three Brevo rules matter for an MSP:
- One DMARC record only. During automatic setup, Brevo asks whether to replace an existing DMARC record. Decline and authenticate manually, or Brevo's
p=nonemay overwrite a client'sp=reject. - Brevo wants its
ruaaddress present. If the record lacks aruatag, Brevo says to appendrua=mailto:rua@dmarc.brevo.com. For several destinations, combine them in oneruatag separated by commas. - Enforced domains must be authenticated first. Brevo states that domains with
p=quarantineorp=rejectmust be authenticated before you can create a sender. Brevo's advice to start atp=noneapplies to new domains, not to clients you have already enforced.
Adding rua@dmarc.brevo.com sends a copy of the client's aggregate reports to Brevo. Brevo publishes a wildcard external reporting authorisation at *._report._dmarc.dmarc.brevo.com with the value v=DMARC1, so receivers will deliver them. Mention this to the client: it shares report data with a third party. Authorising external DMARC report destinations explains the mechanism.
How do Brevo's sender rules tie in with Google and Yahoo?
Brevo frames its requirements around the Gmail and Yahoo rules in force since 1 February 2024 and Microsoft's similar rules announced in May 2025. Brevo asks senders to use a custom domain, authenticate it with DKIM and DMARC, keep spam complaints below 0.3%, and, for bulk senders on a dedicated IP, align the sending domain with the authenticated domain.
Yahoo's sender best practices require bulk senders to publish DMARC with at least p=none, align SPF or DKIM with the From domain (relaxed alignment is fine) and keep the spam rate below 0.3%. Google sets comparable rules in its email sender guidelines. Brevo says it handles PTR records, RFC 5322 formatting and one-click unsubscribe (RFC 8058) automatically.
Watch one alignment detail from Brevo's own examples. If example.co.uk is authenticated, a sender at news.example.co.uk is shown as not compliant, and vice versa. Authenticate each sending domain or subdomain in Brevo separately. The wider picture is in Google and Yahoo sender requirements.
What happens if a client sends from Gmail or an unauthenticated domain?
Brevo replaces the From address. Brevo's example rewrites mycompany@gmail.com to mycompany@5000001.brevosend.com for marketing and to an address on t-sender-sib.com for transactional mail. Brevo calls this a stopgap: recipients may not recognise the address. Free-mail domains such as gmail.com cannot be authenticated at all.
How do you verify a Brevo setup with dig and email headers?
Check DNS first, then a real message. DNS confirms the records exist; headers confirm Brevo actually signs with the client's domain. Run these from any workstation, replacing example.co.uk with the client domain:
dig +short TXT example.co.uk | grep brevo-code
dig +short CNAME brevo1._domainkey.example.co.uk
dig +short CNAME brevo2._domainkey.example.co.uk
dig +short TXT brevo1._domainkey.example.co.uk
dig +short TXT _dmarc.example.co.uk
On Windows, Resolve-DnsName -Type CNAME brevo1._domainkey.example.co.uk does the same job. Expect one brevo-code: string, two CNAMEs ending in dkim.brevo.com, a key behind each one, and exactly one v=DMARC1 record. The DKIM validator checks the published key for a given selector.
Then send a Brevo test campaign to a Gmail or Outlook mailbox and open the original message. In the Authentication-Results header, look for:
dkim=pass header.d=example.co.uk header.s=brevo1
spf=pass smtp.mailfrom=...brevosend.com
dmarc=pass header.from=example.co.uk
The selector may read brevo2 after a rotation. If header.d is a Brevo domain instead of the client's, authentication is not complete. Paste the header into the email header analyzer for a readable breakdown.
What are the common Brevo DMARC failure cases?
Most Brevo failures come from DNS formatting, duplicate records or a sender that does not match the authenticated domain. Brevo's troubleshooting page covers most of them; this table maps each symptom to a fix.
| Symptom | Likely cause | Fix |
|---|---|---|
DMARC fails, DKIM d= is a Brevo domain |
Domain not authenticated, or wrong sender domain | Authenticate the exact From domain or subdomain in Brevo |
| Domain shows "Not authenticated" after 48 hours | Host doubled, such as brevo1._domainkey.example.co.uk.example.co.uk |
Enter only brevo1._domainkey, or add a trailing dot |
| DKIM fails on a Cloudflare zone | Proxy or CNAME flattening turned the CNAME into a TXT | Set the records to DNS only and turn off flattening |
| Two DMARC records found | Brevo's record added beside the client's | Merge into one record, combining rua addresses |
Client's p=reject replaced by p=none |
Automatic setup replaced the existing record | Restore the record and add Brevo's rua by hand |
| DKIM TXT record rejected as too long | 2048-bit key exceeds 255 characters per string | Switch to the CNAME option, or split the string |
| Status reverts to "Not authenticated" | Someone deleted "unknown" records | Restore them and document the selectors |
From shows brevosend.com |
Free-mail or unauthenticated sender | Send from an authenticated client domain |
For general record errors, see common DMARC errors and how to fix them.
What should an MSP check before moving a Brevo client to p=quarantine or p=reject?
Confirm that every Brevo account passes through aligned DKIM, that DNS is documented, and that ownership is agreed. Under RFC 9989 there is no percentage ramp: the percentage tag is gone, so the rollout is p=none, then p=quarantine, then p=reject. You can use the new t=y testing flag while you validate.
- Find every Brevo account. Search aggregate reports for SPF domain
brevosend.com. Different departments may run separate accounts. Client discovery covers the method. - Authenticate every sending domain and subdomain, publishing the records yourself through the manual route so the change sits in your DNS log.
- Check DKIM alignment in reports for at least a full campaign cycle, including transactional mail sent through the Brevo API or SMTP relay.
- Keep one SPF and one DMARC record. Remove stale
include:spf.brevo.comentries only after confirming no dedicated IP setup needs them. - Decide on Brevo's
ruaaddress with the client, and keep your own reporting address first. - Consider a branded subdomain for clients that want two aligned signals before
p=reject. - Document the selectors (
brevo1,brevo2, any legacy TXT selector) and thebrevo-coderecord in the client runbook. - Plan offboarding. When the client leaves Brevo, remove the CNAMEs, the
brevo-coderecord and Brevo'sruaaddress.
DMARC AI classifies sending sources from aggregate reports across all client domains, so an unauthenticated Brevo account shows up as a failing source per client. The broader method for vendors like this is in how to handle third-party senders during DMARC projects, and the same pattern applies to Mailchimp and SendGrid. Before enforcing, also read how to move from DMARC monitoring to enforcement safely.
FAQ
Does Brevo require a DMARC record?
Brevo includes a DMARC TXT record in domain authentication and recommends v=DMARC1; p=none; rua=mailto:rua@dmarc.brevo.com. It warns that Microsoft consumer mailboxes may reject or junk mail from domains without DMARC. Gmail and Yahoo also require DMARC for bulk senders, so in practice every Brevo sending domain needs one.
What are the Brevo DKIM selectors?
Current Brevo accounts use two CNAME records at brevo1._domainkey and brevo2._domainkey, pointing to Brevo-hosted keys under dkim.brevo.com. Older accounts may use a single DKIM TXT record instead, often at mail._domainkey. Copy the exact names and values from the Domains page in your Brevo account.
Do I need include:spf.brevo.com in my SPF record?
Not for shared IP sending. Brevo says SPF is not required for domain authentication, and its default return path is brevosend.com, so your SPF record is never evaluated for that mail. SPF matters if you use a dedicated IP or a branded subdomain; follow the records Brevo shows in those cases.
Is Brevo the same as Sendinblue?
Yes. Sendinblue renamed itself Brevo, and older DNS names such as spf.sendinblue.com and sendinblue-site-verification TXT records still appear in client zones. Brevo's current setup uses the brevo-code TXT record and brevo1 and brevo2 DKIM selectors. Check with Brevo before removing any old Sendinblue records from a live domain.
Can a Brevo client use p=reject?
Yes, once reports show every Brevo stream passing DMARC through aligned DKIM, or through SPF with a branded subdomain. Brevo lets you create senders on domains with p=quarantine or p=reject only after the domain is authenticated. Move through p=none and p=quarantine first, without a percentage ramp.