Sending identity

Receiving providers score the sending domain and the sending address, together with the infrastructure behind them. That means every mailbox carries its own history: bounce rate, complaint rate, how much recipients engage. One mailbox behaving badly damages the domain it belongs to and, through it, every other mailbox on that domain.

The practical consequence is that identities are not interchangeable slots. Adding one adds a reputation you now have to build and maintain, which is work — and it is the reason unlimited mailboxes is a worse offer than it sounds.

The arithmetic is simple. Sustainable volume per mailbox is bounded by the daily cap you are willing to defend, which for cold outbound is typically tens rather than hundreds of messages. Divide the volume you actually need by that number and you have the identity count.

Working backwards from a volume target instead — buying twenty mailboxes because the plan allows it — is how teams end up with twenty ramps to run, twenty reputations to watch and no more replies than they had with three.

Many teams send cold outbound from a domain that is not the one carrying their transactional and customer mail. The reasoning is containment: if outbound goes wrong, invoices and password resets still arrive. The cost is that the secondary domain starts with no reputation and has to build one.

Whichever you choose, authenticate every domain you send from properly, and keep an eye on the alignment between the visible From domain and the one that actually passed authentication — a mismatch there is a common, quiet cause of filtering.

SalesShift sends through mailboxes you connect. Each one carries its own weight, daily cap, ramp and health score, and each is scored independently — so a struggling mailbox loses share of the rotation rather than taking the rest of the pool down with it.