Sending cap

There is the provider's limit — the hard number your mail host enforces, above which sends are rejected — and there is your own cap, which should sit well below it. Confusing the two is the common mistake: running at the provider's ceiling means any burst produces rejections, and rejections are a reputation signal.

For cold outbound the number people converge on is tens per mailbox per day, not hundreds. The exact figure matters less than that it is set deliberately and enforced by the sender rather than discovered by the receiver.

A cap is a ceiling on risk, not a quota to fill. Teams that treat it as a target send exactly the cap every day regardless of whether they had that many relevant people to write to, which is the definition of volume-first outbound and the reason their reply rates fall.

The useful practice is the reverse: set the cap conservatively, let the day's genuinely relevant list fall wherever it falls, and only raise the cap when replies rather than sends justify it.

A cap alone is static. It becomes a system when it composes with three other things: a ramp that raises it gradually for a new mailbox, a weight that lets one mailbox carry more of the pool than another, and a health input that reduces effective share when a mailbox is struggling.

Together those four numbers describe exactly what a mailbox may do today, which is the thing an operator actually wants to be able to read off a screen.

Every mailbox in SalesShift has a daily cap, defaulting to 50 and overridable per mailbox, alongside a weight and a ramp. Today's effective ceiling is the smaller of the cap and the ramp's value for the day, and rotation divides by that number rather than by the raw cap.