Sender reputation

How it works

Providers track engagement and complaint signals per sending identity over time. Sudden volume spikes, high bounce rates or complaints erode the score; steady positive engagement rebuilds it slowly.

Why it matters

Reputation is the foundation under every other email metric. You can write the perfect message, but a damaged sending reputation means most recipients never see it.

Which reputation signals can a sender actually see?

Providers do not publish your score, so you work from indirect evidence. Rejection and deferral text in SMTP responses is the most direct: messages refused with a 5.7.x code that mentions policy or reputation tell you the receiver has decided against you, while bursts of 4xx deferrals often mean you are being throttled. Gmail senders can verify their domain in Google Postmaster Tools to see spam-rate data once volume is high enough, and IP owners can register with Microsoft SNDS.

Outside provider tools, the signals are your own: reply rates by mailbox, complaint counts from feedback loops where available, public blocklist lookups, and seed tests. None is conclusive alone. A drop that appears in two or three of them at once is worth acting on.

Why does one campaign reach the inbox at Gmail but spam at Outlook?

Each provider runs its own filtering model on its own data, so reputation is really several scores, one per receiver. Filtering is also partly per recipient: a person who has replied to you before is more likely to see your next message in the inbox than a stranger at the same company.

Segment results by recipient domain before drawing conclusions. In a hypothetical send of 1,000 emails, 600 to Gmail addresses and 400 to Microsoft-hosted ones, 12 replies from the Gmail side and none from the Microsoft side point to a Microsoft-specific placement problem rather than weak copy. Our notes on Gmail and Outlook deliverability cover the differences in how each filters.

How do sending patterns change reputation over time?

Receivers notice shape as well as totals. A mailbox that sends 30 a day for a month and then 300 on a Monday looks like a compromised account. Long silences followed by bursts, dozens of messages to one company's addresses within minutes, and identical bodies across many mailboxes all read as automation. A steady daily count with some randomness in timing reads as a person.

Sender reputation also spans more than one identity: the mailbox, the domain reputation behind it, and the IPs it sends from. Hitting a spam trap harms all three at once. For planning safe daily numbers, see how many cold emails per day are safe.

How do you tell reputation damage from a content problem?

Run a controlled comparison. Send the same plain message to a few seed inboxes from the mailbox you suspect and from a healthy mailbox on a different domain. If the healthy mailbox reaches the inbox and the suspect one does not, the issue sits with the sender. If both land in spam, look at the message: links to a domain with its own poor history, a tracking domain shared with other senders, heavy HTML, or attachments. Swap one element at a time and resend, so each test answers a single question instead of several at once.

Free tools for this

No signup required — they run in your browser.

FAQ

Related terms