Domain reputation: how mailbox providers judge your domain

Domain reputation is the track record mailbox providers keep for your sending domain. Every cold email you send adds to it, good or bad, and it decides whether the next one lands in the inbox, in spam, or gets rejected before it arrives. Providers don't publish a score or a formula, so what follows comes from what their own documentation and bounce messages say.

What feeds domain reputation

Providers don't share their exact weights, and you shouldn't trust anyone who claims to know them. The inputs they do talk about are consistent:

• Spam complaints. Among the strongest signals. Google's sender guidelines say to keep your spam rate in Postmaster Tools under 0.1%, and never let it reach 0.3%.

• Bounces. Mail to addresses that don't exist reads as a sender who doesn't know who they're emailing. A pile of 550 5.1.1 bounces is the classic pattern.

• Authentication. SPF, DKIM and DMARC passing and aligned. Without them, a provider can't even tie the mail to your domain. See SPF, DKIM and DMARC explained.

• Engagement. Mail people open, reply to and keep, versus mail they ignore or delete.

• Volume patterns. Steady is good, spikes are suspicious. A domain that jumps from nothing to hundreds a day looks the way spammers look. A sudden rate limit like 421 4.7.28 is Gmail noticing.

• Content and links. Domains in your links count too: Gmail's own 4.7.28 variants name link domains as a cause.

Each provider judges separately

There's no one score. Gmail, Microsoft and Yahoo each run their own systems on their own data. A domain can be healthy at Gmail and blocked at Microsoft. Check the providers your list actually uses, and read the bounce text, which often names the cause.

How to see it

Google Postmaster Tools (postmaster.google.com) is Gmail's view of a domain you verify as yours. It shows domain reputation, spam rate, authentication results and delivery errors for mail sent to Gmail. It only shows data when your volume to Gmail is high enough, so a small or new domain may see nothing; that is not the same as a clean bill of health.

Microsoft SNDS (Smart Network Data Services, at sendersupport.olc.protection.outlook.com/snds) is worth knowing, but be accurate about what it is: it reports on IP addresses, not domains. It's most useful if you control the IP ranges you send from. If you send through shared Google Workspace or Microsoft 365 infrastructure, you don't.

Your own bounces are the third source, and the only one that works for every provider. A 550 5.7.1 that names reputation or a block list tells you what that provider thinks. Sort them by provider and by sending domain. High email bounce rate? Burned domain or bad list shows how to read them.

How to protect it

1. Separate cold outreach from your main domain. A bad week in outbound should never touch the domain you use for customers and invoices. See cold email infrastructure.

2. Authenticate before the first send. Verify SPF, DKIM and DMARC pass on a test message.

3. Warm up, then ramp in steps. Follow the signals, not a calendar. Email warmup covers what it can and can't do.

4. Verify the list. Remove dead addresses before they bounce. In SendingOS, email verification is free.

5. Keep complaints low. Send to people who plausibly want to hear from you, write plainly, and make leaving easy.

6. Watch daily. Reputation drifts before it breaks. In SendingOS, bounces are sorted into hard, soft or block with the reason and SMTP code, and domain health rules can pause a drifting domain's inboxes before the whole campaign suffers.

How to recover

There's no fixed number of days for recovery, and nobody can promise one. The shape is:

• Stop the cause first. Pause the campaigns that were sending to a bad list or drawing complaints. More sending on a damaged domain digs the hole.

• Fix what burned it. Authentication, list quality, copy, volume. Fix the real cause, or the domain will burn again.

• Keep light, normal activity going so the domain keeps generating good signals.

• Come back in small steps. Return to campaign volume gradually, to your cleanest, best-fit contacts first, and stop climbing if bounces or rate limits return.

• Know when to retire it. If a domain keeps being blocked after you've fixed the cause, it may cost less to start a new one than to keep nursing the old one. Never move a bad list onto the new domain.

A domain rarely burns overnight; it drifts. The earlier you see it, the cheaper the fix.

Reputation signals you can measure yourself

You can't read a provider's score, but you can track the inputs it uses. Per sending domain, per week, work out:

• Bounce rate: bounces divided by emails sent. Say you send 1,000 emails and 20 bounce: that's 2%. (An example with round numbers, not a benchmark.) Split it into hard and soft, because the fixes differ. See hard bounce vs soft bounce.

• Reply rate: replies divided by emails delivered. A falling reply rate on a stable list is often the first sign mail has started landing in spam.

• Block bounces: any 5.7.x codes that name reputation, a block list or authentication, and which provider they came from.

Compare each against your own last few weeks, not against a number from someone else's business. The trend is what tells you a domain is drifting.

Domain reputation vs IP reputation

Providers weigh both. If you send through Google Workspace or Microsoft 365, the IPs are shared and managed by the provider, so your own domain's behavior is most of what's yours to control. If you run your own sending infrastructure, the IP matters more, and that's where tools like SNDS help. In SendingOS, every client workspace sends on its own isolated IPs, so one client's bad list lands on that client's IPs, not another's. Domain behavior is still yours to protect, workspace by workspace.