Warm-up that pauses itself when the numbers say to.

Six weeks of warming, undone by one Wednesday campaign.

You’ve been nursing the new dedicated IP for a month. A shared spreadsheet tracks the day-by-day cap. Marketing knows not to touch it, but a campaign gets scheduled anyway — 40,000 recipients on an IP that was cleared for 6,000 — and by Thursday morning your bounce rate has spiked, Gmail has dropped your domain reputation, and you’re starting the warm-up again from a worse baseline than you began with.

sendrules moves the daily cap out of the spreadsheet and into the send path. When bounce or complaint rates trend the wrong way, the ramp holds itself at the current cap until the numbers recover — the campaign can’t push more than the IP is ready for, and nobody has to know the schedule to keep it safe.

Advance on health. Hold on risk.

Daily caps advance on schedule when bounce and complaint rates stay under configured thresholds. Cross a gate and the ramp holds at the current cap until you clear the hold or the metrics recover. A live projection strip on the dashboard shows the next seven days at the current pace.

Three starting curves, editable.

• New IP — 50 → 500 → 5,000 → 50,000 over 14 days, doubling every second day when green.
• Migrated IP — starts at 25% of the previous provider’s peak and reclaims full volume over 7 days.
• Paused domain — resumes at 10% of last-known peak with a 3-day observation window before advancing.

Where the gates trip.

• Bounce rate above 3.0% (rolling 24h) — hold.
• Complaint rate above 0.3% (rolling 24h) — hold.
• Deferrals above 15% of attempts (rolling 6h) — hold with alert.
• All thresholds are configurable per Grid and per provider.

Per provider, per IP where the provider exposes it.

Ramp-up runs at the provider level for shared-IP providers (SES, SendGrid, Postmark shared pools). For providers that expose per-IP identity (SES dedicated, Mailgun dedicated, custom SMTP), sendrules ramps each IP independently with its own gate state.