Warm-up for high-volume senders: capacity, limits and requirements

High-volume sending fails for three reasons that look identical from the outside: you crossed a published bulk-sender requirement, you hit a per-mailbox quota or rate limit, or your reputation could not support the volume. The first two are quotas and policies with specific error strings, fixable in days. The third takes weeks. Telling them apart is the first job.

Know the requirements you are now subject to

Since February 2024, Google and Yahoo have applied explicit rules to senders of roughly 5,000 or more messages a day to their consumer domains: SPF and DKIM must be set up, DMARC must be published, one-click unsubscribe conforming to RFC 8058 is required on bulk mail, and the Gmail spam rate in Postmaster Tools must stay below 0.3%. Microsoft announced comparable expectations for high-volume senders to its consumer domains in 2025. Treat 0.3% as a cliff and 0.1% as your working target.

These are not reputation guidelines. They are conditions of delivery, and failing one of them affects mail that would otherwise be fine.

Plan capacity backwards from the limits

Re-check these with the vendor before you build a schedule on one; they change. Then do the arithmetic: target daily volume divided by a conservative per-mailbox rate gives the mailbox count you need. Conservative means well under the cap — if Google allows 2,000, plan on a few hundred, because sustained sending near a ceiling is itself unusual behaviour.

Ramp the fleet, not the mailbox

Ten mailboxes at 40 a day is a safer and more resilient shape than two at 200. One bad mailbox is then 10% of capacity rather than half, and you have room to grow without approaching any single quota. But each mailbox needs its own ramp, and mailboxes created in one batch and started on the same morning are a recognisable pattern. Stagger creation over days, stagger first sends, and keep each mailbox's curve independent.

Beyond that, three structural requirements before any volume:

1. Authentication complete and aligned — one SPF record under ten DNS lookups, DKIM signing with d= matching your From: domain, DMARC published with a reporting address you read.
2. A dedicated sending domain or subdomain per stream, so outbound cannot filter your transactional mail.
3. A verified list. At high volume a 2% bounce rate is a lot of invalid recipients, and it is the quickest route to a block.

Dedicated IP: only with steady volume

A dedicated IP isolates your reputation but has to be kept warm. If your volume is lumpy — big sends then quiet weeks — an idle dedicated IP looks new again and you will do worse than in a well-run shared pool. If you do take one, ramp it over four to six weeks, split each day across providers in your list's proportions, and let deferrals set the pace: a persistent 421 4.7.0 from one provider means hold that provider flat while the others continue.

Monitoring at volume

Daily: hard bounce rate per send, complaint rate, new 4xx and 5xx strings, volume against plan. Weekly: per-provider seed placement, Gmail spam rate trend, per-mailbox outliers against the fleet, blocklist state for domains, IPs and link domains. For IPs you own, Microsoft SNDS is the best free per-IP source.

One thing you can no longer monitor: Google Postmaster Tools retired its domain and IP reputation charts, so there is no reputation grade to track. Spam rate plus per-provider placement is the replacement, and it is a better instrument because it is measured on your mail.

What not to do

Do not plan to send at a published cap. Do not solve a quota rejection by buying more domains — it is a reason to add mailboxes or slow down. Do not use BCC or distribution lists to stretch a quota; recipients are counted anyway and the mail filters worse. Do not raise an SES production limit and immediately send at the new ceiling, because the quota changed and your history did not. Do not pause a ramp for a week and resume at the previous volume. And at this scale especially, do not buy engagement: pools of controlled mailboxes manufacturing opens, clicks and replies are what Validity's Heatwave blocklist, launched 3 September 2026, was built to detect. It lists domains rather than IPs, launched with over a million of them, and publishes no thresholds, so no tool — ours included — can promise exemption. Volume ramping and genuine engagement are the defensible parts.

Steady beats peaky, permanently

The habit that keeps high-volume senders deliverable is a flat curve. A consistent 20,000 a day is easier to sustain than 60,000 on Tuesdays, because providers model your normal and treat departures from it as events. Build the schedule around the floor you can always hit, not the ceiling you occasionally want.

MailPilot is built for fleets: unlimited mailboxes on every plan, warm-up over OAuth for Gmail and Outlook or SMTP/IMAP for Amazon SES, SendGrid, Mailgun, Postmark and custom relays, SPF/DKIM/DMARC checks per domain, real per-provider inbox placement measurement, and per-mailbox reputation recorded daily so one bad sender shows against the rest.