Sign in Get up to 500 free credits

Transactional vs marketing email

Transactional email is a message the recipient needs because of something they did or have: a receipt, a password reset, a shipping notice, an account alert. Marketing email promotes a product or service. The difference matters because the rules differ: US law exempts transactional mail from most of CAN-SPAM, and Gmail and Yahoo require one-click unsubscribe only for marketing mail. Mailbox providers also judge each stream by how people react to it, so the two should be sent apart, on separate subdomains and, at volume, separate IPs.

What counts as transactional

The legal view: CAN-SPAM

The FTC's CAN-SPAM compliance guide names five narrow categories of "transactional or relationship" messages. They:

  1. facilitate, complete or confirm a transaction the recipient already agreed to;
  2. give warranty, recall, safety or security information about something the recipient bought;
  3. announce changes to the terms, features or standing of a membership, subscription, account, loan or other ongoing relationship, including regular account balance information;
  4. give information about an employment relationship or employee benefits;
  5. deliver goods or services the recipient is entitled to under a transaction they agreed to.

What decides is the message's primary purpose. Under the FTC guide, a purely transactional message is exempt from most of the law but must still not contain false or misleading routing information. A message that mixes the two is commercial if the subject line suggests an ad, or if the transactional part does not come mainly at the beginning. An account statement that opens with offers is marketing, whatever its subject line. CAN-SPAM is US law, and this guide is not legal advice; check the rules where your recipients live.

The mailbox provider view

Gmail and Yahoo use the same split for their unsubscribe rule. Google's sender guidelines FAQ says one-click unsubscribe is required only for marketing and promotional messages, and gives password resets, reservation confirmations and form submission confirmations as transactional examples. It also says the line varies by industry and regulation, and that recipients, not Google, decide what a message is to them. Yahoo's sender FAQ likewise exempts transactional mail, with order confirmations and password resets as examples.

Neither FAQ names any other exemption for transactional mail, so plan to meet the authentication and spam-rate rules on both streams (Gmail and Yahoo sender requirements).

The industry view

M3AAWG's Sender Best Common Practices (appendix C) defines transactional messages as confirming a transaction or reporting on the status of the relationship, such as a bank alert or a password change notice. Bulk or marketing messages advertise or build the relationship and are not transactional.

Usually transactionalUsually marketing
Order and payment confirmationNewsletter
Password reset, sign-in codeSale or product announcement
Shipping and delivery noticeAbandoned cart reminder with offers
Security alert, terms changeRe-engagement campaign

Why separate the streams

Reputation. Yahoo's sender best practices say each IP and each DKIM domain has its own reputation, and ask senders not to send bulk or marketing mail from the IPs they use for transactional mail and alerts. Postmark notes that broadcast mail gets lower engagement than transactional mail; on a shared stream, those weaker signals weigh on your password resets too.

Speed. Mailgun warns that time-sensitive transactional mail can queue behind a large marketing batch (IP addresses and sending volume). Postmark built separate infrastructure for the two because mailbox providers often throttle bulk sends, and transactional mail should not wait behind them (Postmark: Message Streams).

Steady volume. M3AAWG (section 2.6) notes that combining transactional and marketing mail creates irregular volume, and consistent volume is part of how IP reputation is judged.

One thing separation does not do: Google counts all mail from the same primary domain, subdomains included, toward its bulk sender threshold (FAQ). A marketing subdomain does not keep your transactional domain out of the bulk sender rules.

How to separate them

Subdomains

M3AAWG (section 2.5) recommends a subdomain per purpose, such as one for offers and another for account mail, with a name that fits the traffic. It also recommends keeping the Return-Path, From:, DKIM d= and link domains consistent within a stream:

Marketing                                   Transactional
Return-Path: bounce@news.example.com        Return-Path: bounce@mail.example.com
From: offers@news.example.com               From: receipts@mail.example.com
DKIM-Signature: d=news.example.com          DKIM-Signature: d=mail.example.com

Google's sender guidelines add that each category of message should use the same From: address, and that one message should not mix different types of content.

Separate IPs or IP pools

Google says that if you send from several IPs, each message type should have its own. Amazon SES lets dedicated IP users build one pool for marketing and another for transactional mail (Dedicated IP addresses for Amazon SES), and Twilio SendGrid recommends at least two dedicated IPs, one per stream, once you reach 250,000 messages a month (Dedicated IP addresses). At lower volume, separate subdomains on shared IPs are often enough; the trade-offs are in shared vs dedicated IPs.

DKIM domains and selectors

Yahoo suggests separating mail types by IP or by DKIM domain, so signing each stream with its own subdomain already does part of the job. Within a domain, DKIM selectors exist so that one domain can hold several keys at once, for example one per mail server or per third party that signs for you (RFC 6376, section 3.1). A separate selector for each sending platform lets you rotate or revoke one platform's key without touching the other's (DKIM selectors and key rotation). Check each one with the DKIM checker.

Feedback-ID for Gmail

Gmail's Feedback Loop reports complaint rates by an identifier you put in a Feedback-ID header. Up to three identifiers come first, then a required sender ID of 5 to 15 characters that stays the same across your mail. A mail-type field is one of Google's own examples:

Feedback-ID: spring-sale:marketing:examplecorp

Google says not to use a value that is unique for every message, and the header only counts when the message is DKIM-signed, after the header is added, by a domain verified in Postmaster Tools. More in Google Postmaster Tools and feedback loops.

Transactional mail goes to people because of a relationship they started, so it needs no marketing consent and no CAN-SPAM opt-out. Keep it that way: once promotions dominate a receipt, the FTC's primary-purpose test can make it commercial.

Marketing mail needs permission. M3AAWG (section 4.1) calls explicit consent the primary rule for bulk and marketing mail and says a new purpose, such as a second newsletter, needs its own opt-in, never a pre-ticked box. Section 4.3 treats consent inferred from a purchase as poor practice. An address given at checkout is a basis for the receipt, not for the newsletter.

Unsubscribing differs too. CAN-SPAM requires a working opt-out in commercial mail, honored within 10 business days. Gmail and Yahoo require one-click unsubscribe headers in marketing mail (one-click unsubscribe). M3AAWG (section 4.6) asks senders to say which lists or types of communication an unsubscribe covers. Leaving the newsletter should not stop password resets.

What sign-up verification does for both

Both streams usually start with the same form field. Checking the address there, before the first message, helps each stream differently (sign-up form validation):

Verification has limits. It cannot prove consent, it cannot tell you whether the person typing owns the address, and on a catch_all domain it cannot confirm the mailbox. A confirmation email does that (double opt-in).

Questions

Is a welcome email transactional?

It depends on content. A short message confirming an account or a subscription may fall under the narrow CAN-SPAM categories, though the FTC gives no welcome-email example; one that mostly promotes products is likely commercial under the primary-purpose test. When in doubt, treat it as marketing and include an unsubscribe link.

Should transactional mail have a List-Unsubscribe header?

Gmail and Yahoo do not require one there. Google notes that recipients decide what a message is, so for borderline mail such as account digests, offering one gives readers a way out other than the spam button.

Do I need a separate domain, not just a subdomain?

No. M3AAWG (section 2.5) recommends subdomains of your main domain and advises against lookalike domains for separating parts of a business.

Try it

Check a sign-up address with the free email checker, or wire the check into your form with the email verification API.

Sources

How we wrote this: drafted by the email-validation.org team with the help of AI writing tools, from our product documentation and the sources linked in the text.

Get up to 500 free credits. No card needed.

New accounts get 500 free credits (100 when you sign up with a free email provider such as Gmail, Outlook or Yahoo, or another personal mailbox such as a paid one like Fastmail or one from your internet provider). Try it on your own list before you buy.

Create your account Open the free email checker

Updated · Maintained by the email-validation.org team