Seed testing measures inbox placement by sending email to pre-created accounts at major providers (Gmail, Outlook, Yahoo) and checking where messages land. It's the most practical method for measuring actual placement, but has limitations: seed accounts lack engagement history, so results may not match real recipient experience. Combine with Postmaster Tools data.
Seed Testing for Inbox Placement: Methodology and Tools
How Seed Testing Works
- A testing service maintains seed accounts at Gmail (multiple accounts), Outlook, Yahoo, AOL, iCloud, and regional providers
- You send your email to these seed addresses, either as part of a real campaign or standalone test
- The service checks each account via IMAP/API and reports where the message landed
- You get a placement report: inbox, spam, tabs, or missing, for each provider
This is the closest thing to knowing your real inbox placement rate. Your ESP's "delivered" metric only means a server accepted the message — it doesn't distinguish inbox from spam.
The Methodology
Test Setup
Include seed addresses in real campaigns whenever possible. Sending a separate test email to seed addresses may get different treatment than a mass campaign because ISPs evaluate sending patterns differently for single messages vs bulk sends.
Use your actual sending infrastructure — same ESP, same IP, same domain, same authentication. Sending from a different system gives you data about that system, not yours.
Test at realistic volume — if you normally send 50K emails, testing with just 70 seed addresses means ISPs process your seeds differently than bulk mail. Include seeds alongside your normal volume.
What to Measure
| Metric | How to Calculate | Target |
|---|---|---|
| Inbox rate | Seeds in inbox / total seeds delivered | > 90% |
| Spam rate | Seeds in spam / total seeds delivered | < 5% |
| Missing rate | Seeds not delivered / total seeds sent | < 3% |
| Tab placement (Gmail) | Seeds in Promotions / Gmail seeds | Track trend |
Practitioner note: The "missing" metric is the one people overlook. If 5% of your seed emails don't arrive at all — not in inbox, not in spam, nowhere — that's a silent deliverability problem that your ESP will never report. It usually means the receiving server rejected the message at the SMTP level.
Limitations of Seed Testing
Engagement Bias
Gmail's filtering heavily weights individual user engagement. A seed account that never opens your emails will get different placement than a subscriber who opens every email. This means:
- Gmail seed results may be pessimistic (seed accounts show more spam) for senders with engaged audiences
- Gmail seed results may be optimistic for senders with unengaged audiences
Sample Size
With 2-5 seed accounts per provider, you're working with tiny sample sizes. One account's placement can swing your "Gmail inbox rate" from 100% to 50%. Look at trends across multiple tests, not individual results.
Behavior Differences
Seed accounts don't click, reply, move to folders, or report spam. They're passive recipients. This means seed testing can't measure the impact of your engagement optimization strategies.
Practitioner note: I tell clients to think of seed testing like a weather report — it's directionally correct and useful for planning, but your specific experience may differ. A "70% inbox rate" from seed testing doesn't mean exactly 70% of your real subscribers see your email in their inbox. It means there's likely a placement problem worth investigating.
Best Practices
Frequency
| Scenario | Testing Frequency |
|---|---|
| Normal operations | Weekly automated tests |
| Pre-major campaign | Test with campaign content 24-48 hours before |
| Post-infrastructure change | Immediately, then daily for 1 week |
| During warmup | Every 2-3 days |
| After deliverability incident | Daily until resolved |
Integration with Other Data
Seed testing works best when combined with:
- Google Postmaster Tools — real Gmail reputation and spam rate data
- Microsoft SNDS — Outlook-specific IP reputation
- ESP engagement metrics — open/click rates from real subscribers
- Feedback loops — complaint data from ISPs
No single data source gives you the full picture. Seed testing fills the "actual placement" gap that other tools miss.
Seed List Hygiene
Keep your seed list clean:
- Remove any seed addresses that have been deactivated
- Don't let seeds accumulate in suppression lists (they shouldn't bounce or complain)
- Segment seeds from real subscribers in your ESP reporting so they don't skew metrics
Practitioner note: One common mistake: clients add GlockApps seeds to their list and then their ESP's engagement metrics look weird because 70 addresses never open anything. Create a separate segment or tag for seed addresses and exclude them from engagement calculations.
Tools
GlockApps — 70+ seed accounts, $59-149/month, best for mid-market
InboxReady (Validity) — Combines seed testing with panel data for more accurate Gmail results. Enterprise pricing.
Manual testing — Create your own free accounts at Gmail, Outlook, and Yahoo. Less convenient, no automation, but free. Works for occasional spot-checks.
If you need help setting up seed testing and interpreting results alongside your other deliverability data, schedule a consulting call — I'll build a testing framework matched to your sending profile.
Sources
- Validity: Inbox Placement Testing Methodology
- GlockApps: How Seed Testing Works
- M3AAWG: Best Practices for Measuring Email Delivery
- Google: Postmaster Tools FAQ
v1.0 · April 2026
Frequently Asked Questions
What is seed testing in email?
Seed testing sends your email to maintained 'seed' accounts at every major mailbox provider. The testing service then checks whether the email arrived in the inbox, spam, promotions tab, or wasn't delivered. This gives you real placement data across providers.
How accurate is seed testing?
Directionally accurate but not perfect. Seed accounts don't have real engagement history with your brand, so engagement-based filters (especially Gmail) may treat them differently than actual subscribers. Use as one data point alongside other metrics.
How many seed accounts do I need?
Testing services typically maintain 2-5 seed accounts per provider. More accounts per provider increases accuracy but also increases noise. GlockApps uses about 70+ total seed accounts across all providers.
Should I add seed addresses to my regular sends?
Yes, for ongoing monitoring. Add the seed list to your regular campaigns so you're testing under real sending conditions. Just ensure the seed addresses don't skew your engagement metrics in your ESP.
Can seed testing detect promotions tab placement?
Yes. Gmail seed accounts can report whether your email landed in Primary, Promotions, Social, or Updates tabs. This is the only reliable way to check tab placement at scale.
Want this handled for you?
Free 30-minute strategy call. Walk away with a plan either way.