Seed Panel (Diagnostic)

James's real, aged personal inboxes across Gmail, Outlook, iCloud, and Yahoo serve a specific, bounded role in a Capture migration: they are a placement diagnostic first, and a light one-time positive nudge on the first one to three sends second. That is the full extent of their value. They are not a reputation engine. Real member replies on Day 1 will dwarf anything the seed panel contributes — understand that framing before deploying it.

The seed panel is real but marginal. Its primary job is to catch a spam placement problem before 500 real members see it. Its secondary job is a handful of genuine early signals on the first send or two. After that, step back and let real engagement drive the subdomain's reputation.

Why It Is Still Worth Using

Two things the seed panel does that nothing else can:

Placement catch before the real list. A placement failure visible to a small panel of seed inboxes is recoverable in an afternoon. One visible to 500 members is a trust problem that takes weeks to repair. The diagnostic catches misconfigurations — broken DKIM, a DMARC policy gap, a content-level trigger — before they reach real members.

The only practical per-domain Outlook read. Microsoft's SNDS (Smart Network Data Services) operates at the IP level on the shared Mailgun pool. It tells you nothing about how mail.clubname.co.uk is performing specifically. The only way to see per-domain Outlook placement for your subdomain is to hold a real Outlook inbox and check it manually. That is the seed panel's practical, irreplaceable function for Outlook monitoring.

What It Cannot Do

Be explicit with clubs and internally about what the seed panel does not do:

It does not replace clean, consented member data. High bounces and stale addresses undo it entirely.
It does not replace authentication. SPF, DKIM, DMARC, alignment, and the custom tracking CNAME must all pass before the diagnostic send. The seed panel cannot compensate for authentication failures.
It does not replace the engagement-gated ramp. Sending to the most-engaged Tier A members first, earning genuine signals from them, and widening incrementally is where real subdomain reputation is built.
It does not guarantee inbox placement. Placement is influenced by content, list quality, complaint history, IP behaviour, and dozens of other signals. The seed panel is one light input among many.

Inbox Requirements

The panel must meet all of the following criteria. An inbox that fails any one of them should not be used.

Criterion	Requirement
Account age	Real personal accounts used normally over years — not recently created
Provider spread	At least one inbox per major provider: Gmail, Outlook (Hotmail/Live), iCloud, Yahoo
Activity	Actively receives and sends real email — dormant inboxes behave anomalously
Ownership	James-owned or directly agency-controlled — must be confident they will behave naturally
Count	James's existing small panel of real, aged inboxes — never create new accounts to expand it

Automatic disqualifiers:

Accounts created within the last twelve months with little inbox activity
Accounts opened specifically for this purpose in bulk
Accounts accessed via scripts, bots, or automation tools of any kind

The Hard Line: Genuine Versus Manufactured

This is the section most likely to go wrong. Every mailbox provider runs detection systems that identify artificial engagement — and those systems have become significantly more sophisticated as AI-driven warming tools have proliferated.

Real aged inboxes behaving naturally = acceptable.
Manufactured engagement at scale = forbidden.

The test for any individual action: would a real person, who genuinely wants to receive this club's emails, actually do this?

Action	Acceptable?
Open and read the email	Yes
Click a link you are genuinely curious about	Yes
Reply to ask about an upcoming event	Yes
Add the sender to contacts because you want it in your inbox	Yes
Mark Not Spam because it landed in spam and you want it there	Yes — but only if it actually landed in spam
Create fifty Gmail accounts last Tuesday from the same IP	No
Use a script to auto-open, auto-click, auto-move 200 emails in ten minutes	No
Run a third-party inbox placement or seeding service	No
Have every seed inbox open, click, and mark-safe within a twenty-minute window	No
Apply the seed panel on an ongoing cadence beyond the first three sends	No — dormant seed inboxes on a permanent list can trigger a dead-list penalty

Why the ongoing cadence is forbidden specifically: AI warming detection systems now identify inboxes that open and engage at unnaturally consistent rates relative to list-wide behaviour. A panel of inboxes that keeps engaging when the rest of the list has gone quiet is a detectable pattern. Spamhaus CSS (Content Spam Source) targets senders who use artificial engagement infrastructure; a listing there means mail is blocked across a significant proportion of commercial email filtering worldwide. Recovery takes weeks. This is not a theoretical risk — it has caught real senders using third-party seeding services.

The line is genuineness combined with no automation, not the actions themselves.

Pre-Launch Sequence

Run in order. Do not reorder or skip steps.

Step 0 — Intelligent Golf farewell (T minus seven to ten days)

Before a single seed inbox receives anything, send a final broadcast from the club's existing Intelligent Golf identity. This is the single highest-leverage deliverability action in the entire migration — it rides IG's already-trusted sending infrastructure to prime real member inboxes before the new domain ever appears in their folder.

The message content:

From [date], our emails will arrive from a new address at mail.[clubdomain].co.uk. To make sure you keep receiving our updates, please add this address to your contacts and check your junk folder if you do not see our next email.

Mirror it on the member-portal login page if the club can update it. Clubhouse or pro-shop signage is a worthwhile secondary nudge. Do not skip this step.

Step 1 — Authenticate fully (T minus three days)

The true gate for any send is full authentication — not volume. Verify every item before touching the seed panel.

If any item fails, resolve it before proceeding. Sending to real members with broken authentication is the fastest route to damaged domain reputation.

Step 2 — Diagnostic send to seed panel (T minus two days)

Compose a test email in GHL that closely resembles the real first campaign. Use the same template structure, link density, and image approach planned for the Trust Bridge send. Do not use a plain "ping" test — you want to simulate the real send as closely as possible.

Send to all seed inboxes. Wait at least one hour before checking. Over a four to eight hour window, check each inbox individually and record results:

Inbox	Provider	Placement	Action taken
Example: inbox1@gmail.com	Gmail	Primary inbox	Open + click
Example: inbox2@outlook.com	Outlook	Focused inbox	Open + reply
Example: inbox3@icloud.com	iCloud	Junk	Mark Not Spam, open

Placement decision gate:

Outcome	Decision
All land in inbox or Promotions	Proceed to Step 3
Any seed inbox lands in Junk/Spam	Stop. Diagnose root cause — authentication gap, content issue, or blocklist hit — before any member send. Re-run diagnostic after fixing. Do not proceed to real members.

Step 3 — Seed panel positive-signal actions (T minus one to two days, first sends only)

This step applies to the first one to three sends only. After that, real member engagement takes over entirely.

Seed inboxes that received the email in inbox or Promotions perform their assigned natural behaviours, staggered over four to six hours. Vary behaviour across the panel — no two inboxes should do identical things on the same cadence:

Behaviour	Approximate proportion	Example action
Open only	30%	Open, read for ten to fifteen seconds, close
Open and click a genuine link	20%	Click one real link — club website, event page
Open and reply naturally	15%	Short genuine reply — "Thanks, looking forward to it"
Open, click, add sender to contacts	15%	Open, click, add `info@mail.[clubdomain].co.uk` to address book
Open and reply with a question	10%	"Will there be a summer social this year?"
Do not open	10%	Leave unopened — real lists always have non-openers

Critical rule: mark Not Spam only where the email genuinely landed in Spam. Pre-emptive "not spam" actions on inbox-delivered mail do nothing and create an anomalous signal.

Allow twenty-four hours after seed actions before the first real send.

Step 4 — First real send (Trust Bridge to Tier A)

The seed panel's primary role is now complete. On launch day, send the Trust Bridge to Tier A only — the most-engaged members (clicked, replied, booked, or transacted in Intelligent Golf in the last 30 days — never opens; non-role-based addresses, committee members, recent joiners, active event participants).

GHL has no native send throttling — it bursts the entire queue against shared-pool rate limits (approximately 300/hour on Mailgun Flex by default). Manually batch and cap per-hour. Do not one-shot the queue. Monitor bounce and complaint rates before expanding volume.

Keep five to ten seed inboxes on the live send list for ongoing placement monitoring. Remove the rest — they have done their job.

Step 5 — Engagement-gated ramp

Proceed into the standard engagement-gated ramp (default seven to fourteen days, widening from Tier A through Tiers B–C (Tier D on the 60-day fallback only) in approximately 1.5× steps, gated on Postmaster spam rate below 0.10% and hard bounce below 0.50%). The thirty/sixty-day schedule is the documented fallback for genuinely messy data clubs — not the default.

Metrics: What to Watch

Judge the seed diagnostic — and the full ramp — on signals that are not distorted by Apple Mail Privacy Protection. Opens are recorded automatically by MPP for roughly 49% of UK email users, making them unreliable as an engagement signal.

Signal	Green	Amber	Red
Gmail spam complaint rate	Below 0.10%	0.10–0.20%	Above 0.20% urgent / ≥0.30% hard fail
Hard bounce rate	Below 0.5%	0.5–2%	Above 2% pause / above 3% stop and revalidate
Unsubscribe rate	Below 0.30%	0.30–1.0%	Above 1.0% review
Auth pass rate (SPF/DKIM/DMARC)	~100%	Any unexplained drop	—

Source Gmail spam rate from Google Postmaster Tools (not GHL in-platform). Source bounce, deferral, and block data from Mailgun analytics — GHL's in-platform view on custom SMTP shows only opens and clicks, not delivery failures. Gmail retired its domain and IP reputation dashboard on 30 September 2025; Postmaster Tools spam-rate Pass/Fail is now the authoritative Google read.

Quick-Reference Checklists

Pre-launch diagnostic

Seed panel positive-signal actions (first one to three sends only)

Actions staggered over four to six hour window
Behaviour mix varied — no two inboxes performing identical actions on the same cadence
Not Spam marked only where email genuinely landed in Spam
Replies sent naturally, short, and not identical across inboxes
Add-to-contacts completed for designated inboxes
Twenty-four hour wait after seed actions before first real send

First real send

Tier A segment identified: most-engaged members (clicked, replied, booked, or transacted in IG in the last 30 days — never opens)
Role-based addresses (info@, office@, admin@) suppressed; club-stakeholder roles verified individually
GHL manual batching configured — no one-shot queue blast; cap at approximately 300/hour
Google Postmaster Tools and Mailgun analytics open and monitored before expanding to Tier B
Five to ten seed inboxes retained on live list for ongoing placement monitoring; remainder removed

The full pre-cutover sequence — Step 0 priming, authentication, list segmentation, Trust Bridge format guidance, and the engagement-gated ramp schedule — is in the Operator Runbook. Volume gates and advancement criteria are in the Ramp Schedule.

Seed Panel (Diagnostic)

On this page