EmailOctopus SMTP Settings and Configuration (2026)

If you landed here searching for “EmailOctopus SMTP settings”, you almost certainly want one of two things, and this page covers both. Either you need the sending-domain DNS records EmailOctopus generates when you verify a domain — that is the real settings screen — or you need to fire transactional mail such as password resets, receipts and WordPress notifications, which EmailOctopus cannot do at any price. For that second job you pair EmailOctopus with a real relay: Amazon SES is the one EmailOctopus itself names, while Postmark, SMTP2GO and Resend are the usual turnkey answers. If you want one account that does both, look at Brevo or Sender.net.

Quick reference

Why EmailOctopus has no SMTP relay

EmailOctopus is a marketing-only ESP by design. Its product surface is campaigns, automations, forms and landing pages, all composed inside the dashboard, and its billing model is list size rather than message volume. A relay is the opposite shape of product: it charges per message and needs abuse tooling for arbitrary payloads. EmailOctopus never built that side, so there is no transactional product, no message-submission endpoint and no submission service on ports 587 or 465 to point a client at.

The history is what makes people expect otherwise. EmailOctopus launched in 2015 as a thin front end over your own Amazon SES account — the product still sold today as EmailOctopus Connect. In August 2019 it removed the AWS requirement for new signups, citing that over 60% of them abandoned at the Amazon setup step, and moved the main product onto its own partnered infrastructure. That infrastructure was never exposed as a relay, so there are now two sending backends behind one dashboard and neither is an SMTP service you can authenticate against.

What you actually need instead

Sending transactional email

Password resets, receipts, invoices and one-to-one system mail are out of scope entirely, and no plan or add-on changes that. Knowledge base article 199 points you at Amazon SES — a separate service you configure and pay for yourself. Choose on operational appetite: Amazon SES is cheapest per message and shares DNS ground with Connect, but you own sandbox exit, quota increases and reputation. Postmark is fastest to a working relay. SMTP2GO and Mailgun sit between the two.

Connecting a CMS or app

WP Mail SMTP, and framework mailers in Laravel, Django or Rails, all expect a host, a port and a credential pair. EmailOctopus supplies none of them. Configure the plugin against your chosen relay for outbound system mail, then connect EmailOctopus separately for list building via the official WordPress plugin or the v2 API. The two paths never touch: one carries mail over SMTP, the other pushes contacts over HTTPS.

Routing app and monitoring notifications

Alerting, cron output, form-handler notifications and CI mail are the classic “just give me an SMTP host” case, and the traffic you least want mixed with marketing sending. Point them at a relay on a dedicated subdomain such as mail.yourdomain.com, so a noisy monitoring stream never touches the domain your campaigns sign with. EmailOctopus stays on newsletters, where its $0 Starter plan covers 2,500 subscribers and 10,000 emails per month.

Sending domain authentication

This is the screen most people mean by “SMTP settings”. EmailOctopus does not authenticate you with an SMTP username and password — it authenticates your domain with DNS records, and the records differ between the main product and Connect.

SPF and DKIM setup

On the main EmailOctopus product, open Senders, click Start verification, and the app returns four CNAME records plus one TXT record. The values are account-specific and never printed in the knowledge base, so copy them exactly as the app shows them — and do not append your domain if your DNS host already does. EmailOctopus provisions a custom MAIL FROM subdomain in the eom.yourdomain.com form and handles the SPF record on its own side, so there is no include: string to paste into your SPF TXT record.

; EmailOctopus (main product) — Senders › Start verification
; Shape only. The four CNAME hosts and targets are generated per account
; and shown once, inside the app. Copy them verbatim from that screen.

TYPE    HOST                       VALUE
CNAME   <host-1 from app>          <value-1 from app>
CNAME   <host-2 from app>          <value-2 from app>
CNAME   <host-3 from app>          <value-3 from app>
CNAME   <host-4 from app>          <value-4 from app>
TXT     <host-5 from app>          <value-5 from app>

; SPF: nothing to add. EmailOctopus publishes SPF on its own
; MAIL FROM subdomain, eom.yourdomain.com.

On Connect the same job moves to the AWS console, because the mail leaves from your own SES account. Create the identity under Amazon SES › Verified identities, add the DKIM CNAMEs SES generates, and optionally configure a custom MAIL FROM domain with the MX and TXT records SES supplies. Nothing about that setup is EmailOctopus-specific; it is standard Amazon SES domain verification, and it stays valid if you later move that SES account to another tool.

DMARC alignment

EmailOctopus documents a starting policy of a TXT record at _dmarc.yourdomain.com with the value below, and recommends a free report aggregator for a few weeks before tightening to quarantine or reject. It flags DMARC as required practice above 10,000 emails a day.

; DMARC — start permissive, then tighten
Host:  _dmarc.yourdomain.com
Type:  TXT
Value: v=DMARC1; p=none;

; Add an rua= address once you have an aggregator configured, then
; move p=none to p=quarantine and finally p=reject.

Alignment is where a pairing setup goes wrong. DMARC passes when SPF or DKIM aligns with the visible From domain: EmailOctopus aligns via its MAIL FROM subdomain and DKIM CNAMEs, your relay via its own records. Both can pass on the same organisational domain at once, provided each vendor’s records are actually present. A single _dmarc TXT record covers the domain and its subdomains, so publish exactly one and never duplicate it per vendor.

Sending limits

Read the X-RateLimiting-Remaining header on every API response rather than guessing at the bucket. Note what is not in the table: no hourly send ceiling for the main product, no recipients-per-message cap and no maximum message size — the last because attachments are not supported at all. Treat those as undocumented rather than unlimited.

Common mistakes

Pairing EmailOctopus with an SMTP provider

The recommended setup

Run EmailOctopus for campaigns and one relay for everything a machine sends. Verify your root domain in EmailOctopus with the four CNAMEs and the TXT record, verify a dedicated subdomain — usually mail.yourdomain.com — with your relay, and point your application, CMS and monitoring at that relay on port 587 with STARTTLS. Splitting the signing domains keeps a bounced marketing send from dragging down password-reset delivery, and lets you read the two reputations independently.

If you are already on EmailOctopus Connect, Amazon SES is the natural relay because the identity is already verified in your AWS account: generate SES SMTP credentials in the console and you have a working host without touching DNS again. If you would rather not run AWS plumbing, Postmark or SMTP2GO are the quickest routes. Marketers who want one vendor for both jobs should compare Brevo and Sender.net, which ship a relay alongside campaigns, or Mailchimp, where the relay is a paid add-on. MailerLite sits in the same marketing-only category as EmailOctopus and will not solve this either.

Keeping the two systems from conflicting

Three failure modes account for nearly every broken pairing. The first is double-sending: if both systems react to the same event — a Zapier automation adding a buyer to a list that fires a welcome email, while your app sends its own order confirmation — the customer gets two messages minutes apart. Decide per message type which system owns it, and silence the other.

The second is DMARC drift. One _dmarc TXT record governs the organisational domain, so tightening to p=reject because your relay is aligned will also reject EmailOctopus mail whose CNAMEs were added incompletely. Confirm both vendors pass in aggregate reports first, over at least two weeks of data.

The third is suppression divergence. A relay’s bounce and complaint suppression list is entirely separate from the EmailOctopus one. An address that hard-bounced on transactional mail stays perfectly mailable in your EmailOctopus list, and vice versa. Nothing syncs them — use the v2 API and your relay webhooks to reconcile on a schedule, or you will keep mailing addresses your own application knows are dead.

Frequently asked questions