Inbox Placement Testing: How to Run a Seed Test and Read the Results

An inbox placement test sends a real message to a set of seed mailboxes at different providers, then reports where each copy landed: the inbox, a tab such as Promotions, the spam folder, or nowhere. It answers one question: did this message, from this domain and sending setup, reach the inbox at these providers today? The procedure is short. Send your exact production message to the seed addresses through your real sending setup, wait a few minutes, then read the placement and authentication results per provider.

It measures inbox placement rate, which no ESP dashboard can give you, because an ESP only sees that a receiving server accepted the mail.

When a Seed Test Is Worth Running

Run one in these situations:

• Before a major campaign, especially to a segment you have not mailed recently.
• After any infrastructure change: a new sending domain, a new dedicated IP, or a move to a different ESP. For a new domain or IP, test during email warmup or IP warming so you see problems at low volume.
• After a drop in opens or a spike in complaints, to separate a placement problem from a content or audience problem.
• On a schedule. Iterable’s guidance is that “performing one test per sending infrastructure/mail stream once every week or two is usually sufficient” (Iterable). Twilio warns that you “should avoid over-using seed testing” because seed accounts never open or click (Twilio SendGrid).

How to Run an Inbox Placement Test, Step by Step

1. Choose your seed source. A vendor gives you a managed seed list and a report. Building your own means creating test accounts at Gmail, Outlook.com, Yahoo and any provider your list leans on, then checking each by hand. Vendor lists save manual checking and add authentication results. Your own accounts are free and match your audience, but take upkeep and pick up history if you ever open or reply. Iterable notes that reusing test addresses for other checks means “you’re engaging with the emails and, therefore, creating a history with the sending infrastructure.”
2. Send the production message, unchanged. Use the same From domain, the same ESP and IP pool, the same subject, body, links and headers as the real send. GMass says it plainly for its free tester: “Don’t change the content just for the test” (GMass). A test of a cleaned-up copy tells you nothing about the real one.
3. Add seeds without skewing the real send. Send to the seeds as their own small batch through the same stream, rather than adding dozens of never-engaging addresses to your main list. GlockApps suggests that if you cannot send to its list at once, you “send the email to groups of 25 addresses at a time,” and that repeated tests of one message be spaced “30-60 minutes” apart (GlockApps).
4. Wait for results. Most tools report within minutes. A seed that has not received anything after that is usually a delivery failure, not a slow filter: GlockApps says a missing result is “usually because the message did not reach those seed addresses.”
5. Capture the authentication results. Record SPF, DKIM and DMARC pass or fail for the test message. GlockApps reports these alongside placement. SPF, DKIM, and DMARC are required by Google for bulk senders: “Set up DMARC email authentication for your sending domain” (Google sender guidelines).
6. Save the full report with the date and what changed. Twilio notes “trend analysis often provides more value than a single test result.”

How to Read the Results

Read placement per provider, not as one blended number. This table gives the usual meaning of each outcome and the first thing to check.

Result	What it usually means	Check first
Inbox	The filter accepted the message from this sender at this moment	Whether the result holds across providers and across two runs
Promotions or other tab	Treated as promotional, not as a rejection	Whether that is acceptable for this message type; for transactional mail it is a content and format signal
Spam	The provider’s filter distrusts the sender, the content or the links	Authentication results, then sender reputation and any blocklist listing
Missing	The message never reached the seed	Bounce and block events in your ESP logs, then SMTP responses and blocklists
Mixed (some seeds inbox, some spam at one provider)	The provider is unsure about you	Repeat the test; a shaky reputation tends to show as inconsistent results

A worked example (illustrative)

Suppose a test of a product-update email from news.example.com lands in the inbox at Gmail and Yahoo but in spam at Outlook.com. Authentication shows SPF and DKIM passing and DMARC passing at all three. Because authentication is identical everywhere, the difference is the provider’s own filtering, which points to reputation or content as seen by Microsoft. Next check blocklist status, then the links in the message.

Now weight it by your audience. If most of your list is on Microsoft-hosted addresses, that spam result is your biggest problem even though two of three providers passed. If Outlook.com is a sliver of the list, fix it without pausing the campaign. Provider mix decides what a bad result costs you.

What a Seed Test Cannot Show

A seed test measures how a provider treats a message sent to accounts with no engagement history. It does not show how that provider treats your actual subscribers.

• No engagement signal. Seed accounts do not open, click or reply, so filtering that depends on recipient behavior is not reproduced. Twilio says “seed accounts do not behave like real humans,” and GlockApps states that it “does not simulate user engagement on these addresses.” Twilio also warns that “good seed placement does not guarantee strong inboxing for disengaged subscribers,” and that a seed landing in spam “does not necessarily mean your real audience is seeing the same issue.”
• It is a sample, not a forecast. GlockApps describes its own result as “not a prediction of your whole audience,” and notes that “provider coverage is weighted toward US consumer providers.” With a few seeds per provider, one misplaced copy swings the percentage, so look at the pattern, not the number.
• One send, one setup. Iterable notes seed testing gives feedback “for a particular sending infrastructure,” so a pass on one IP pool says nothing about another.

The counterpart is data from real recipients. Google Postmaster Tools reports spam rate, reputation and authentication for mail sent to personal Gmail accounts, though “data might be missing if the total number of messages for a given day is too low” (Google). It covers Gmail only. Twilio’s advice is to treat seed testing as “one signal among many.”

What to Do After a Bad Result

Work in this order, because earlier items are cheaper and more often the cause:

1. Authentication. Confirm SPF, DKIM and DMARC pass and align for the From domain.
2. Sender reputation and blocklists. Check sender reputation signals and whether your domain or IP is on a blocklist.
3. Content and links. Compare against the patterns in why emails go to spam: link domains, image-heavy layouts, misleading subjects.
4. List quality. Remove unengaged and invalid addresses with email list cleaning; stale lists drive complaints and can include a spam trap.
5. Change one thing, then retest the same message so you know what moved the result.

Seed Testing Tools at a Glance

Details below come from each vendor’s own page, checked on October 11, 2026. Pricing and limits change, so confirm before buying. For a wider comparison see our roundup of deliverability tools.

Tool	What it checks	Free option
GlockApps	Inbox, tab and spam placement; SPF, DKIM, DMARC; blacklist status; spam filter scores	Free plan with 2 spam test credits
Mailgun Optimize	Inbox placement tests plus email validation	First month free; then $49/mo for 25 placement tests
SendForensics	Seed list of major providers worldwide, reporting inbox, spam or a subfolder such as Promotions	No free tier stated; says plans include unlimited placement tests (fair use)
GMass inbox tester	15 Gmail and Google Workspace seed addresses, some running filters such as Barracuda, Mimecast and Sophos	Free

Mail-Tester is a different kind of tool: it analyzes one message sent to a temporary address, and its FAQ does not claim to show placement at Gmail.

Frequently Asked Questions