The standard Instantly playbook in 2026 looks like this: buy 10 to 30 Google Workspace or Microsoft 365 mailboxes across 3 to 5 domains, warm them for two weeks, plug them into a campaign, let rotation spread 30 to 50 sends per mailbox per day. At 30 mailboxes and 40 sends each, you are doing 1,200 cold sends per day. That is substantial volume. It is also fragile volume. One mailbox drifting into Spam drags the whole campaign's reply rate down, and the dashboard tells you everything is fine.
Add the same seed list to every campaign. Because Instantly rotates sender, each seed will eventually receive a copy from every mailbox in your rotation pool. Look at the seed inbox grouped by the From header and you instantly see which specific mailbox is landing in Junk. Rotate it out for re-warm before it poisons the rest.
Why Instantly rotation needs per-inbox seeding
Instantly's multi-inbox rotation is its biggest feature and its biggest hidden risk. Every mailbox in the pool has its own reputation history, its own warm-up curve, its own IP routing inside Google or Microsoft. They are not interchangeable. Provider filters build a per-sender reputation model. One mailbox that sends a 20% open-rate copy four days in a row will start getting downgraded at Outlook's SmartScreen even if the other 29 mailboxes in the pool are perfect.
The dashboard shows you aggregate sends, aggregate opens, aggregate replies. If one mailbox drops from 40% open rate to 8%, the pool average still looks healthy. You do not notice until the whole campaign drifts. At that point you have wasted weeks of warm-up across multiple mailboxes chasing a problem that started with one.
Per-inbox seed setup inside Instantly
Instantly does not offer a dedicated "send a seed from each mailbox" feature. The workaround is to exploit rotation itself.
- Generate 20+ seeds with the free Inbox Check tool.
- In Instantly, open the campaign, go to Leads and import the seed list as regular leads.
- Leave rotation on. Set sends-per-mailbox-per-day high enough that in a typical day each mailbox will hit at least one seed address. With 30 mailboxes and 20 seeds, one day of sending gives you a roughly full matrix.
- After 24 to 48 hours, open each seed inbox and group received messages by the
Fromheader. Note which folder each landed in. - Build a mailbox x provider matrix. Any row (mailbox) that shows Junk / Spam at two or more providers is burned.
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.
Reading the mailbox-by-provider matrix
A mailbox matrix looks like this after one full rotation cycle:
Mailbox Gmail Outlook Yahoo Mail.ru
alice@domain1 Inbox Focused Inbox Inbox
bob@domain1 Inbox Junk Inbox Inbox
carol@domain2 Promo Focused Inbox Spam
dave@domain2 Inbox Focused Inbox Inbox
...bob@domain1 is drifting at Outlook. Pull it out of rotation and back into warm-up only. carol@domain2 is showing two problems — Promotions at Gmail and Spam at Mail.ru — which typically means content on that specific mailbox's send path triggered SpamAssassin rules. Re-seed carol once after removing the tracking pixel, see if it recovers.
How often to run the per-inbox seed test
Before scaling above 500 sends/day: run the matrix weekly. Above 1,000 sends/day: twice a week. Right after adding new mailboxes to the rotation pool: within 24 hours of those mailboxes going live. After any DNS change (DKIM rotation, SPF update): within the same day. After switching the campaign copy: the day the new copy goes live.
Costs nothing if you use free seeds. Costs the reputation of the whole pool if you skip it.
Warm-up score vs seed placement
Instantly's warm-up score is genuinely useful — it reflects how well the mailbox is interacting with the internal warm-up network. But the warm-up network is not Gmail, and it is not Outlook. Two mailboxes with identical 100% warm-up scores can produce completely different seed placement results because the signals that matter at Google Postmaster (volume growth curve, user complaint rate, authentication consistency) are not the same signals as the warm-up network's internal opens and replies.
The rule we use: warm-up score is necessary but not sufficient. 100% warm-up + Outlook Junk in seed results = still a problem. Below 70% warm-up + Outlook Inbox in seed results = warm up longer, do not scale volume yet.
Scaling from 100 to 1000 sends/day
The staircase we see work:
- Day 0 to 14: warm-up only. No cold sends. Seed test once at day 10 as a control — you want Gmail Inbox, Outlook Focused.
- Day 14 to 21: cold sends at 10/mailbox/day. Seed test at day 17.
- Day 21 to 30: ramp to 25/mailbox/day. Seed test at day 24 and day 28. Remove any mailbox showing two-provider Spam.
- Day 30 to 45: full volume, 40/mailbox/day. Weekly seed matrix.
- Day 45+: steady state. Twice-weekly seed matrix. Add new mailboxes in batches of 5, warm them separately, only merge into the active pool after a clean solo seed test.
Microsoft 365 tenants sometimes ship with the sending mailbox accidentally on an IP shared with a compromised neighbour. SPF and DKIM pass, warm-up looks fine, but Outlook consistently lands Junk. You cannot fix this from the Instantly side — you replace the mailbox. Per-inbox seed tests catch it on day one instead of day fourteen.
Authentication checks on the seed side
When you open a seed, do not just look at the folder. Open the full headers and read Authentication-Results. You need:
spf=passwith the sender domain (not Instantly's envelope domain).dkim=passwith a selector you recognise from your DNS.dmarc=pass— alignment is what actually matters in 2026.
If any of those three come back fail or none from one mailbox but pass from the others, the problem is a DNS issue at that mailbox's domain. Fix before you scale, not after.