Two different things hide under the word "warmup", and the debate only makes sense once they are separated. Volume ramping — starting a new domain or inbox at a handful of sends and increasing slowly — is uncontroversial; every mailbox provider tells new senders to do it. Automated warmup networks — pools of bot-run mailboxes that open, rescue and reply to each other's mail — are the contested part. Nearly all the evidence that warmup networks work comes from companies selling them; nearly all the evidence that they are dangerous is anecdotal or comes from email platforms that do not sell them.
Ramp volume: yes, always. Automated warmup pools: they probably help a brand-new inbox clear the first weeks, they are not a substitute for wanted mail, and their downside risk is rising. Nobody has published an independent controlled study that settles it.
The case for warmup networks
- New inboxes have no history. A mailbox that sends 30 messages to strangers on day one looks like a compromised account. Every vendor and most practitioners prescribe 2–4 weeks of ramp for both Google and Microsoft (EmailBison, a sequencer's blog).
- Engagement is a ranking signal. Gmail and Outlook use replies, "not spam" clicks and moves to inbox as positive signals (see How filtering works and Reputation and blocklists). Warmup pools manufacture exactly those signals.
- Vendor-reported lifts. Instantly claims "30% higher inbox placement" for warmup users. A practitioner blog, OneAway, reports its own Q1 2026 test: a manually warmed test domain reached 79% inbox placement after 28 days versus 23% without warmup. Neither publishes methodology.
- The market votes with its wallet. Pre-warmed inboxes sell at double the price of fresh ones — Instantly $10 vs $5, Smartlead $9 vs $4.50. Agencies pay that premium to skip the ramp, which says they believe the ramp matters.
The case against
- It is fake engagement, and filters are built to find fakes. Resend's September 30, 2026 post argues: "If you rely on automation to simulate interest, inbox systems eventually detect that pattern. When that happens, filtering tightens, reputation declines, and the domain becomes harder to recover." Resend is a transactional ESP — it does not sell warmup, but it does have an interest in clean shared infrastructure.
- A named detection service. The same post says Validity's "Heatwave" identifies domains with synthetic warming patterns, shares the data with Comcast, Proofpoint, Spamhaus and SURBL, and that listings are permanent.
- Coordinated patterns are visible. OneAway claims "Gmail can now identify when 1,000+ mailboxes are all replying to each other with similar timing patterns" and that domains using "safe volume ramps" saw 34% higher placement than those using fake-engagement tools across 47 domains. No source is given for the Gmail claim.
- Infra vendors list it as a ban trigger. ColdSire names "rapid warmup that looks suspicious" among the reasons Google suspends Workspace accounts.
- The signals are cheap to discount. A reply from a known warmup-pool address is worth nothing if the filter knows the address is in a pool. The value of a pool falls as its membership becomes identifiable.
Validity "Heatwave" could not be verified: Validity's own page returned 404 to this pass and no second source was found. Treat the claims about data-sharing with Spamhaus and permanent listing as unconfirmed until Validity documents them.
What Google and Microsoft have said
Neither has published a rule that names warmup tools. Google's bulk-sender rules (Google and Yahoo sender rules) and Microsoft's May 2025 Outlook rules (Microsoft rules: Outlook.com and Microsoft 365 limits) are about authentication, unsubscribe and complaint rates. Both providers' terms prohibit deceptive or abusive use in general language (Google Workspace AUP). The autumn 2025 Google suspension wave was attributed by vendors to account sourcing (EDU and nonprofit "panels") and shared infrastructure, not to warmup specifically (Primeforge; Salesforge newsletter).
There is no public statement from Gmail, Outlook or Yahoo about warmup networks, and no independent, controlled study comparing pools against plain ramping. Everything on this page is vendor-run or anecdotal. This is one of the Open questions.
Where the evidence lands
| Claim | Evidence quality | Read |
|---|---|---|
| New inboxes need a volume ramp | Strong (consistent across providers and practitioners) | True |
| Warmup pools improve placement for new inboxes | Weak (vendor claims, no methods) | Plausible, short-term |
| Warmup pools keep working on mature inboxes | Very weak | Unproven; reply quality matters more |
| Filters detect and punish pools | Weak-to-moderate (one named detector, unverified; anecdotes) | Rising risk |
| Pre-warmed inboxes are worth ~2x | Market price, not proof | Buyers believe it |
The practical consensus among volume senders is to warm for two to three weeks, keep a low level of warmup running alongside campaigns, and cap campaigns at 15–30 sends per inbox per day (Scaledmail caps; Puzzleinbox, a vendor blog, recommends 15–20). That rule of thumb works partly because of warmup and partly because low per-inbox volume is itself the protection — see Volume math: what 10k and 100k emails a day cost.
The volume model's real deliverability lever is not warmup; it is spreading sends thinly across many inboxes. Warmup is what makes a fresh inbox usable in week three instead of week eight.
What this means for an entrant
- Include warmup, default it to conservative, and let customers switch it off. Buyers expect it (see Built-in warmup); a sceptical segment will pay for a sequencer that does not force it.
- Separate ramp from engagement fakery in the product. A ramp engine that schedules low-volume real sends to the best-fit prospects first is defensible in any future policy climate; a bot pool may not be.
- Make measurement the headline. Inbox-placement tests, Postmaster Tools and SNDS ingestion, and reply-rate tracking per inbox give customers evidence instead of faith. See Inbox placement testing.
- Do not build a warmup network as your wedge. It is free in every competitor, needs scale to be credible, and carries the most policy risk in the stack. See Warmup networks.
- Watch Heatwave. If Validity or Spamhaus confirm listings for warmup participation, every bundled pool becomes a liability overnight — and a sequencer without one becomes the safe choice.