Inbox Placement Testing and Seed Lists

Your email platform reports a delivery rate. That number tells you how many messages were accepted by receiving servers. It does not tell you where they were placed. A message in the spam folder counts as delivered. Inbox placement testing exists to answer the question delivery rate cannot: did the email reach the inbox?

Delivery Rate vs Inbox Placement

• Delivery rate is the share of sent messages that receiving servers accepted.
• Inbox placement rate is the share of delivered messages that landed in the inbox instead of spam.

A sender can have a 99% delivery rate and very poor inbox placement. Mailbox providers do not notify senders when mail is filtered to spam, so placement has to be estimated.

How Seed List Testing Works

A seed list is a set of test mailboxes maintained by a testing service at many mailbox providers, such as Gmail, Outlook, Yahoo and regional services.

The process is simple:

1. The testing tool gives you a list of seed addresses and an identifier.
2. You send your real campaign, from your real sending setup, to those addresses.
3. The tool logs in to each seed mailbox and records where the message arrived.
4. You receive a report by provider: inbox, spam, a tab such as Promotions, or missing.

Most reports also include authentication results, spam filter scores and blocklist checks for your sending IP and domain.

What Seed Tests Are Good For

• Checking a new domain, IP or email platform before sending at volume
• Comparing two versions of a template, subject line or sending domain
• Spotting a sudden change at one mailbox provider
• Confirming that a fix has worked
• Verifying authentication and header setup from the recipient's side

Used this way, as a comparison tool and an early warning, seed testing is valuable.

The Limits of Seed Testing

Seed results are an estimate. Treat them with some caution.

Seed Mailboxes Are Not Real Subscribers

Modern filtering is personalized. Gmail and Microsoft weigh how each recipient has interacted with a sender. Seed mailboxes have no history with you. They never open, reply or move your mail, so they show how a provider treats your mail for a recipient with no relationship to you. Engaged subscribers often see better placement than seeds suggest. Unengaged ones may see worse.

Small Samples

A seed list may hold a few dozen mailboxes per provider. A swing from 80% to 60% at one provider may reflect only a handful of mailboxes.

Seeds Can Be Identified

If seed addresses are sent a message separately from the main campaign, they may be treated differently. For the most realistic result, include the seeds in the actual send.

Business Mail Is Hard to Model

Corporate recipients sit behind security gateways and company-specific rules. No seed list can represent every organization's filtering.

Panel Data

Some services measure placement with a panel of real users who have agreed to share data about their mailboxes. Panel data reflects genuine engagement history, which makes it more representative for consumer mail. Panels are smaller and less controllable than seed lists, and coverage varies by provider and country. Large senders often use both.

Other Ways to Measure Placement

Seed tests should sit alongside signals from your real audience.

• Google Postmaster Tools shows spam complaint rate, authentication results and delivery errors for mail sent to Gmail.
• Microsoft SNDS shows data for mail sent to Outlook.com from your IP addresses.
• Engagement by mailbox provider is one of the most useful views. If click rates at one provider fall sharply while others hold, placement at that provider has probably dropped.
• Bounce and deferral logs often show trouble before placement changes.

Be careful with open rates. Apple Mail Privacy Protection and some security tools load images automatically, which inflates opens. Clicks and replies are more reliable.

A Practical Testing Routine

1. Run a seed test before the first send from any new domain, IP or platform.
2. Establish a baseline with a few tests of your normal campaigns.
3. Test again whenever something significant changes: template, sending domain, volume or provider.
4. Investigate when engagement at one mailbox provider drops.
5. Test regularly if you send at high volume, so that you learn what normal looks like.

Change one thing at a time. If you alter the template, the subject line and the sending domain together, a test cannot tell you which change mattered.

Reading a Poor Result

When a test shows spam placement, work through the likely causes in order:

• Authentication. Did SPF, DKIM and DMARC pass and align?
• Reputation. Is the spam complaint rate elevated? Is the IP or domain blocklisted?
• Scope. Is the problem at every provider, or only one?
• Content. Does a plain-text version of the same message place better? Check links and tracking domains.
• Recent changes. What is different from the last send that placed well?

A problem at a single provider usually points to reputation with that provider. A problem everywhere usually points to authentication or a blocklist.

Key Point

Seed tests show how mailbox providers treat your mail in the absence of engagement. Your real results depend on the relationship you have with actual recipients. Use tests to detect change and compare options, and use engagement and Postmaster data to judge how your program is really performing.