There is no public Google statement endorsing warmup tools. There is no page on developers.google.com or support.google.com that says “to warm up a new sending domain, join a warmup network.” Every warmup vendor's marketing carefully avoids quoting Gmail directly, because Gmail has never given them anything to quote.
What Gmail has published is a series of documents that, read together, describe a system that treats coordinated engagement pools as exactly the kind of pattern it is trying to filter out.
Gmail reputation is computed from complaints, spam-trap hits, filtered-at-delivery rates, authentication alignment, and engagement diversity. A warmup pool produces synthetic engagement with low diversity, and filters are explicitly tuned to detect low-diversity engagement patterns. Translation: warmup pools generate the exact signal Google wants to devalue.
What Gmail publishes about reputation
Three public surfaces describe Gmail's sender model:
- The Gmail sender guidelines — authentication, complaint rate below 0.3%, one-click unsubscribe, no alignment failures.
- Postmaster Tools dashboards — domain reputation (four buckets), IP reputation, spam rate, authenticated traffic, encryption, delivery errors.
- The 2024 sender requirements announcement — bulk senders above 5,000/day must pass DMARC alignment, keep complaints below 0.3%, and support RFC 8058 one-click unsubscribe.
None of these mention warmup, engagement pools, opens, or synthetic replies. The omission is not accidental.
What “reputation” really is
Gmail's reputation signal is a rolling classifier output, not a human-curated score. It takes in:
- Complaint rate from the “Report Spam” button — the single heaviest input.
- Filtered-at-delivery rate — messages placed in Spam by Gmail's content filter before the user sees them.
- Authentication alignment — does SPF/DKIM/DMARC match the From domain, consistently.
- Spam-trap hits — mail to addresses Google operates as traps. A single hit can crater an IP for weeks.
- Engagement — replies, forwarding, marking as important, adding to contacts. Not opens.
- User behaviour patterns — reads that end in immediate delete vs. reads that end in reply; add-to-contacts vs. unsubscribe.
What a warmup pool looks like from Gmail's side
Imagine you work on the spam team at Google. You have access to all inbound Gmail traffic and all outbound Gmail-to-Gmail traffic between corporate/workspace accounts. You run anomaly detection over this corpus. What does a 5,000-domain warmup pool look like to you?
- Large cluster of domains that reply to each other's mail at near-100% rate and never reply to anyone outside the cluster.
- Thread contents that follow a small number of templates, detectable by simple n-gram fingerprinting.
- Reply latency distributions that are tighter than organic email (bots reply at 3 minutes, humans at 3 hours).
- “Mark as important” and “add to contacts” actions that cluster on the same domain pairs repeatedly.
- Diurnal patterns that don't match the claimed geography of the participating domains (a US-based domain with 24/7 even reply distribution is obviously automated).
Every one of those is a textbook coordinated-inauthentic-behaviour signal. The same kind of signal Google's search team uses to detect link farms, their YouTube team uses to detect view bots, and their ads team uses to detect click fraud.
Why Google doesn't publicly ban warmup
Because they don't need to. Filtering is graduated, invisible, and recoverable for legitimate senders. The path Google prefers is:
- Detect the coordinated pattern.
- Stop counting the synthetic engagement as positive reputation signal.
- Let the pool members' real-world complaint rate, spam-trap hits and filtered-at-delivery rate do the talking.
This is why so many warmup dashboards show “95% inbox” while real outbound from the same domain lands in Promotions or Spam. Google didn't ban the pool — they just stopped counting it. The pool members are still patting themselves on the back in an empty room.
The engagement-diversity signal
One of the least-discussed pieces of Gmail's reputation model is diversity. Engagement from 10,000 unique domains over 90 days is worth much more than the same engagement volume from 100 repeat domains. Warmup pools fail this test by design — they are the same 500 – 5,000 domains every day.
A side-effect: a real customer newsletter list of even 2,000 organic subscribers produces more reputation signal than any warmup pool of 5,000 participants, because the diversity is orders of magnitude higher per message.
Register your sending domain in Gmail Postmaster Tools. That's Google showing you their view, not your warmup dashboard's view. Combine it with an Inbox Check placement test for the outcome a real recipient sees.
What to do instead of a pool
- Authenticate completely. SPF, DKIM, DMARC aligned. DKIM key ≥ 1024 bits.
- Earn diversity. Any real opt-in traffic, however small, beats pool volume.
- Keep complaints below 0.1%. Gmail's public threshold is 0.3% but 0.1% is where high-reputation senders live.
- Watch Postmaster weekly. If domain reputation moves down, stop sending and diagnose. Do not double down on warmup to compensate.
- Measure from outside. Independent seed mailboxes are the only honest placement signal.