Omnisend pitches itself as the e-commerce marketing automation platform that does email, SMS, and web push in one flow. For DTC brands that is a real advantage — a cart abandoner gets an SMS at 30 minutes, an email at 4 hours, a push notification at 24 hours, all orchestrated from one workflow.
When something in that flow fails, though, it is almost always the email leg. SMS delivery is a binary — it either reaches the phone or returns a carrier error that Omnisend logs and surfaces. Push notifications are similar. But email can be "delivered" (server accepted it) and still sit in Promotions or Spam for hours, invisible to the customer. That is where seed addresses pay for themselves.
For every Omnisend automation with an email node, attach seeds to the audience. Seeds receive email (and you can skip SMS steps for them by filtering on phone = empty). Read per-step placement in Inbox Check to find the step that silently drops to spam.
Why the email leg fails most in a multi-channel flow
Three structural reasons the email step is the weak link in an Omnisend flow:
- Cadence confusion. When a flow alternates SMS and email, the email step may fire at odd times (e.g. within thirty minutes of the previous SMS). Gmail sees a cluster of messages to the same recipient over a short window and scores it as aggressive bulk.
- Template inheritance. Omnisend email templates share a header/footer across the flow. If the footer has a broken unsubscribe link or a List-Unsubscribe header misconfiguration, every email in the flow fails authentication in the same way. Seed tests on one step reveal the issue on all of them.
- Shared-IP drift. Lower Omnisend tiers share sending IPs. A neighbor's reputation drop hits you. The symptom looks like an internal problem but is external. Only seed placement data surfaces it.
Setting up seeds inside Omnisend
Omnisend uses the term Contact for everything (SMS and email subscribers). Seeds need to be email-only contacts so they do not trigger SMS routing.
- Generate a fresh seed batch from the free Inbox Check tool — 20+ addresses across Gmail, Outlook, Yahoo, Mail.ru, Yandex, ProtonMail, GMX and more.
- In Omnisend, go to Audience → Import contacts. Set the CSV header for phone to empty (or omit the column).
- Add a custom property
contact_type = seed. This keeps seeds visible and filterable without confusing them with your real customers. - Create a Segment:
contact_type = seed. This is the audience you enroll in every flow for testing. - Critically, build an inverse segment
contact_type != seedfor your real audience sends and all revenue reports.
The free Inbox Check tool generates 20+ fresh seed addresses per test across Gmail, Outlook, Yahoo, Mail.ru, Yandex, ProtonMail and more. No signup, no credit card.
Skipping SMS steps for seeds
Because seeds have no phone numbers, Omnisend handles this natively most of the time — SMS steps simply do not fire for contacts with empty phone fields. But some flows are configured with a "must have phone" exit condition, which means seeds get kicked out of the flow at the first SMS branch.
Two options:
- Configure the flow to branch, not exit. Use Omnisend's split-on-channel node: if phone is present, send SMS; if empty, skip to the next email step. This is the correct flow design anyway, since real customers sometimes unsubscribe from SMS while keeping email.
- Give seeds a dummy phone number. Use a test sandbox number (e.g. a Twilio test SID) or a non-routable number that Omnisend will attempt and silently fail on. Not ideal — you lose the signal that SMS delivery would report.
Option 1 is strongly preferred. It doubles as better flow design for real customers and works cleanly for seeds.
Reading placement per step of an Omnisend flow
A classic DTC cart abandonment flow in Omnisend looks like:
- Trigger: cart abandoned > 30 min
- Step 1: SMS (or skip if no phone) — "Come back"
- Wait 4 hours
- Step 2: Email — "You left something behind"
- Wait 20 hours
- Step 3: Email — "Still interested? Here's 10% off"
- Wait 48 hours
- Step 4: Push notification
- Step 5: Email — "Last call — your cart expires"
Seeds enrolled in this flow will receive three emails (Step 2, Step 3, Step 5) over roughly three days. Inbox Check reports each one separately. The pattern you are watching for:
- Step 2 lands inbox, Step 3 lands Promotions, Step 5 lands Spam. This is the decay pattern — every step with the coupon/discount keyword drops further. Fix Step 3 and Step 5 subjects and body content.
- All three land inbox but spam score rises. You are on the edge. Small drift on any step (neighbor reputation, list hygiene) will push you over. Clean proactively.
- Step 5 lands Spam at Mail.ru/Yandex only. Russian-audience-specific issue. Usually DMARC alignment or an untranslated English subject line.
Measurement over time
One-off tests catch bugs. Continuous testing catches decay. The highest-value pattern for Omnisend users is weekly or bi-weekly seed tests on the top three revenue-driving flows (welcome, cart abandonment, post-purchase). Track two metrics per flow:
- Average inbox placement across seed providers on the critical email step (usually the coupon email).
- Auth pass rate (SPF + DKIM + DMARC = 3/3).
Plot these week over week. A gradual decline means your content or your domain reputation is drifting. An abrupt drop means a DNS change, a template change, or a shared-IP event. The time between the drop and your first customer complaint is typically two weeks — the seed data lets you catch and fix within days.
Omnisend lets you A/B test subject lines on campaign sends but not easily on automation emails. Use seed tests to simulate an A/B on an automation: draft two versions of the coupon email, send each to a different half of the seed batch, compare placement and spam score. The winning version goes live.