Definition
A seed address is an inbox or test record that the business controls and deliberately includes in a send or a workflow, so that the team can inspect exactly what a recipient would receive. Three features separate it from casually checking the email yourself.
- It travels the real route. The message reaches the seed through the sending platform, the domain's signing and the provider's filtering, so the seed's copy carries the full headers, the final links and the placement that provider chose.
- It is controlled and permanent. The business owns the inbox or the record, keeps it from send to send and knows what it should receive, so a difference stands out.
- It is marked. A seed is recognisable at a glance and excluded from the numbers, so testing never becomes part of the results.
Seed addresses for email and seed contacts for workflows
Email guides use the word for inboxes. A few addresses the business monitors at each major mailbox provider let a team read a delivered message's full headers and see whether it landed in the inbox, a tab or junk at that provider.
Automation guides talk about seed contacts. The guide to automation governance runs them through every branch of a workflow before it is switched on and after any material change, then reads what each one received. They are permanent fixtures, named so nobody could mistake them for a customer and excluded from reporting and from any sync to sales. The same contacts serve the monthly spot check that suppression and consent conditions still hold.
The two meanings share a purpose: put a known recipient where real recipients are about to be, and look before they do.
In practice: authentication, fixes and test data
- Checking authentication before a platform carries real mail. The guide to choosing an email platform has trial accounts send to seed addresses at the main providers and read the headers for which domain signed the message, whether it aligns with the From address, and the unsubscribe headers, so the platform is judged on what it actually sends.
- Testing a fix. When a sent email pointed somewhere wrong, the guide to fixing a mistake in a sent email tests the repair by clicking the link inside the delivered copy at a seed address rather than visiting the corrected page directly, because readers arrive through the link, not the address bar.
- Keeping test data out of the numbers. The guide to test environments and test data asks every tool's test records to follow one written marking convention, with a small permanent set rather than fresh records for each test. Seeds are the email and automation case of that rule.
What seeds cannot do is measure placement. The deliverability guide is explicit that seeds are not stand-ins for real subscribers: they do not open, click or complain the way a list does, and providers weigh that behaviour. A seed in the inbox is a smoke test passed; engagement by mailbox provider over time remains the measure.
A yarn shop's first send from a new platform
Gilvoard Yarns, an online wool and knitting supplies shop, moved its 9,400 subscribers to a new email platform. Before migrating the list it set up 4 seed addresses, one at each mailbox provider that held at least 5% of its subscribers, and sent them its first campaign. Three copies arrived in the inbox; the fourth was filed as junk. Reading the full headers on all 4 showed the same fault: the platform was still signing with its own shared domain, so the message was not aligned with the shop's From address. The owner set up signing on the shop's domain, waited for the records to update and resent to every seed. All 4 headers now showed aligned passes and all 4 copies reached the inbox. Only then was the list moved, and engagement by provider over the next 6 weeks, not the seeds, showed placement holding.
Related terms
DeliverabilityWhether sent email reaches the inbox rather than spam or rejection, decided by sender reputation. is what seed addresses glimpse but cannot measure: whether a sender's mail reaches the inbox across its real list. Email AuthenticationRecords and signatures, SPF, DKIM and DMARC, that let mailbox providers check that a message comes from the domain it claims. is what their headers are most often read for, since signing and alignment show up in a delivered message rather than in a settings screen. A Staging EnvironmentA private copy of a website or tool where changes are tested before release, kept behind access control and out of search and production analytics. is where changes are tested away from customers; seeds do a similar job inside live systems, where a separate copy is not available. The Suppression ListA list of addresses or contacts excluded from every send, such as unsubscribes, hard bounces and spam complaints. is one of the conditions seed contacts are run against, to confirm that people who should never be mailed are still excluded.
Common confusions
- A seed list for a lookalike audience. Advertising uses "seed list" for the source audience a lookalike audience is built from. That is an audience, not a test recipient.
- Inbox placement at the seed means inbox placement for the list. Seeds do not engage, so a clean result is a smoke test. Placement for real subscribers is read from engagement by provider.
- Testing on a personal inbox is the same thing. A colleague's own address is not permanent, not marked and often not at the providers the list uses.
- Seed contacts can be created fresh for each test. Throwaway records multiply, escape filters and leak into reporting and sales syncs. A small permanent set is easier to exclude and compare.
