The tools that told cold senders how they were doing have been switched off. Google retired the Postmaster Tools v1 domain and IP reputation dashboards on 30 September 2025. Microsoft's June 2026 SNDS overhaul stripped complaint samples and removed trap-hit counts from 22 July 2026, and SNDS is IP-based anyway, so it says nothing about senders on Google or Microsoft mailboxes. Microsoft now says the SCL header "no longer holds the same meaning in cloud organizations". Google's sender rules do not even apply to mail sent to Workspace accounts, which is where B2B cold email goes. See Reputation and blocklists.
Senders are flying blind. The sequencers answer with campaign-level pause buttons and a warmup pool.
The opening
The leaders protect the campaign, not the mailbox:
| Vendor | Bounce protection (as of Oct 2026) | Scope |
|---|---|---|
| Instantly | 5% default, checked after 200 emails (help) | Campaign |
| Smartlead | User-set threshold; rate = bounces over total leads in the campaign (help) | Campaign |
| Woodpecker | 2% default, shown after 20 emails (help) | Campaign |
| PlusVibe | Account automations on 3-day bounce rate, warmup health, reply rate → cut limits, recovery period (help) | Mailbox |
Smartlead's denominator is the giveaway. Dividing by total leads means a 10,000-lead upload hides an early bounce spike. A mailbox burning in campaign A keeps sending in campaign B. See Bounce protection.
Instantly is moving: its Deliverability AI Agent (Hyper Growth and above) pauses campaigns and replaces risky accounts, and its automated placement tests can pause a mailbox when placement falls below a threshold. Both rest on seed tests and are sold as tiers.
The pain is in the reviews. Smartlead users report a "domain blacklisted after warmup" (Sep 2026) and a system that "works in the morning, but within hours it completely fails" (Jan 2026). Smartlead's own help centre has an article titled "69 reasons for campaigns getting paused". The autumn 2025 Google wave hit an estimated ~25% of reputable infra providers' clients. That figure is a vendor estimate, and nobody had a mailbox-level early warning. See Provider crackdowns.
Meanwhile the warmup pool, the industry's default answer, is under attack. Resend calls warmup fake engagement that filters learn to detect (Sep 2026). Instantly sells a "premium" pool at $500/month per workspace, which admits that pool quality varies. See Does warmup work?.
Resend's claim that Validity's "Heatwave" identifies warmup traffic and shares it with Spamhaus and others could not be confirmed; the Validity page returned 404. Treat it as a risk to watch, not a fact.
Who pays and how much
Nobody buys "health" as a line item today. They buy proxies for it:
- Placement testing: Smartlead SmartDelivery $49 / $174 / $599 a month; Instantly placement tests on its plans (Instantly).
- Standalone warmup and monitoring: MailReach $19.50 per mailbox per month, which comes to $1,560/month for 80 mailboxes.
- Premium warmup pools: Instantly $500–1,000/month.
The value shows up as retention, not as a SKU. A pre-warmed inbox sells for about $4.50–5 a month more than a fresh one (Instantly, Smartlead). That premium is the market price of a mailbox's health, and every burnt domain throws it away.
Why incumbents have not closed it
- Their architecture is campaign-first. Limits, rotation and pausing were designed around campaigns and inbox counts, and "unlimited inboxes" made inbox-level care an afterthought (Inbox rotation).
- Warmup is their moat and their story. Instantly claims a network of "over 1,000,000+" accounts. An engine that says "measure live sends, don't fake engagement" undercuts their marketing.
- Placement is an upsell. Selling seed tests as a separate tier makes "included, and inferred from your real sends" a pricing weapon (Inbox placement testing).
What you would build first
- Mailbox and domain as first-class objects with history: sends, bounces (parsed RFC 3464 status codes), deferrals, reply rate and complaint signals per destination provider.
- Rolling-window bounce maths. Rate over emails sent in the last N sends or hours, per mailbox and per domain. First check at about 50 sends, not 200.
- Adaptive limits. Each mailbox's daily cap moves with its own signals. The published "safe" volumes disagree too much to hard-code: 80–100/day (Instantly's 2022 blog), 20/day hard cap on AirMail, 25 Google / 10 Microsoft (ScaledMail). See Sending limits and throttling.
- Rotate out, ramp back. Pull a mailbox at the narrowest scope (mailbox, then domain, then list), keep it alive with low real volume, and ramp it back on evidence.
- Inferred reputation with honest confidence. Label every estimate "high", "low" or "insufficient data". Google now "doesn't track open rates", and Apple MPP touches roughly 55–60% of opens, so build on replies and response codes, not opens.
- Placement from live sends, with optional seed tests from a partner such as EmailGuard or MailReach.
How the leaders would respond
Instantly is already building toward this with its Deliverability Agent and will likely move protection to the mailbox within a year. Smartlead will fix its formula once someone embarrasses it publicly. They have far more send data than an entrant, so the inference models they could train are better. Your edge is honesty and defaults: on by default, on every plan, with the reasoning shown.
Scores, argued
Pain: 4. Burnt domains and suspended mailboxes are the most expensive thing that happens to a sender, and the 2025 Google wave showed the scale. Not a 5, because most senders blame their copy or list before their tooling.
Gap: 3. Campaign-level protection is universal and PlusVibe already does mailbox-level rules. Instantly's agent is closing in. Inferred reputation with stated confidence and placement inferred from live sends are what nobody ships.
Size: 3. As a standalone monitoring product the wallet is MailReach-sized. Inside a sender it protects the whole account's revenue, which is why it scores above 2.
Moat: 2. The logic is easy to copy and incumbents have more training data. Only an entrant's cross-customer telemetry and credibility with deliverability-literate agencies resist copying.
Speed: 3. You need live sending volume to tune it. An overlay that reads Smartlead and Instantly events is possible but limited, because you cannot throttle their mailboxes from outside.
Safety: 4. It reduces the customer's provider risk. The residual risk is that adaptive limits fail to catch a "wave" pattern, as everyone's did in 2025.
What would kill it
Google ships the promised new Postmaster dashboards with per-domain reputation for low-volume senders, and Instantly makes mailbox-level protection free on Growth. Then first-party data plus a free incumbent feature leaves nothing to sell. Neither has happened as of October 2026.
What this means for an entrant
- Make it the heart of your sender, not a product. Every other opening that sends email (Agency operating system, EU-native compliant outbound, Governed volume for sales teams, Sending engine as an API) depends on it. Without it you churn on deliverability within a quarter. See Sending architecture.
- Stack it with Transparent infrastructure. The health engine decides when to rotate a mailbox out. Infra with multi-provider failover decides where the domain goes next. Together they are "suspension insurance".
- Do not build a warmup network. Ship a conservative real-send ramp, make bot warmup optional, and market "no fake engagement". See Built-in warmup.
- Add European destinations. No vendor documents routing or seeds for GMX/WEB.DE, T-Online, IONOS or OVH (ESP matching, by absence). If you lead with EU-native compliant outbound, those are your customers' recipients.
- Publish your bounce maths. A one-page methodology beats a green dashboard with agencies who have been burned.