Sender reputation: how mailbox providers score you

Sender reputation is the running assessment a mailbox provider keeps of your sending identity: how likely mail from your domain and IP addresses is to be wanted by its users. Every placement decision consults it. Authentication proves who you are; reputation records how you have behaved, and it is built entirely from evidence you generate: complaint rates, engagement patterns, bounce rates, trap hits, volume consistency, and protocol hygiene.

Domain reputation vs IP reputation

Reputation attaches to two identities at once. Domain reputation follows your authenticated domain, the DKIM d= domain or SPF domain, wherever the mail is physically sent from, which is what makes it portable across ESP migrations. IP reputation follows the sending addresses, which is why new IPs need warming and why shared pools mean shared fate. Providers weight them differently: Gmail leans on domain identity, Microsoft weights IPs more heavily, and both consult both. The practical consequence: your domain carries your history, your IPs carry your infrastructure choices, and both need to be clean.

What feeds the score

The inputs, roughly in order of weight at consumer providers: user-reported spam complaints against delivered volume; per-recipient engagement behaviour (reads, replies, deletes without reading); hard bounce and unknown-user rates, which measure list quality; spam trap hits, which measure acquisition and hygiene sins; volume patterns, where sudden spikes read as compromise; and authentication consistency, since streams that intermittently fail SPF or DKIM look unstable. None of these is a metric you optimize directly. Each is a byproduct of who you mail and how they respond.

Where you can actually see it

Gmail Postmaster Tools shows domain and IP reputation in four bands plus the user-reported spam rate. Microsoft SNDS shows per-IP filter verdicts in color bands plus trap hits. Yahoo exposes no dashboard; its complaint feedback loop is the visibility. Everything else is inference from your own telemetry: deferral rates by provider, per-provider engagement trends, and seed placement. The dashboards lag one to three days, so your own data detects incidents first and the dashboards confirm them.

How reputation moves

Damage is fast and recovery is slow by design: a single bad campaign can drop a domain band within days, while climbing back takes weeks of consistently clean sending, because providers weight recent negative evidence heavily and rebuild trust gradually. The levers that move it upward are the unglamorous ones: list hygiene that keeps you mailing people who respond, complaint management that removes the unhappy before they report you, stable volume patterns, and flawless authentication. There is no reputation fix that routes around sending wanted mail.