What’s the first step in reputation repair?

The first step in reputation repair is not repair. It is triage. You stop the bleeding before you try to heal.

If your complaint rate is above 0.10% on Gmail or 0.30% on Yahoo, or your bounce rate is over 2%, you are actively making things worse with every send. Google's sender guidelines treat 0.10% as the warning line and 0.30% as the hard ceiling that triggers throttling and spam folder placement. Yahoo's Sender Hub uses similar thresholds. Keep sending into that, and you train Gmail and Yahoo's filters that your domain belongs in spam.

Here is the order I run it in when a client lands in trouble.

1. Pause or throttle, then look at Postmaster

Stop your highest-risk sends first. Promotional blasts to cold or re-engagement segments are usually where the damage is coming from. Transactional and one-to-one mail can keep flowing if it has a separate stream.

Log into Google Postmaster Tools and check Domain Reputation, IP Reputation, Spam Rate, and Authentication. Red or Bad on Domain Reputation means Gmail is already filtering you. Spam Rate above 0.10% means users are hitting the spam button. If you do not have Postmaster set up, that is your real first step. You cannot fix what you cannot see. We cover the setup in how to monitor your domain reputation.

For Microsoft, enroll in SNDS and JMRP for IP-level data and complaint feedback.

2. Find the source, do not guess

Most reputation crashes come from one of four things:

• A bad list import (purchased data, scraped contacts, old re-engagement file)
• A spamtrap hit (recycled addresses on a list you have not validated in 6+ months)
• An authentication break (SPF, DKIM, or DMARC alignment failing after a config change)
• A sudden volume spike (10x your normal send in a day, which providers read as a compromised account)

Pull your last 30 days of sends and segment by list source, campaign, and send date. The campaign where complaints spiked or bounces jumped is your patient zero. Cross-reference with any list you imported, any template change, or any new IP you started using.

Run your sending domain through MXToolbox to check public blocklists like Spamhaus, SORBS, and Barracuda. A listing tells you the damage is already visible outside the mailbox providers.

3. Isolate damaged infrastructure

If one subdomain or IP is the problem, take it out of rotation. Send the rest of your mail from clean streams while you investigate. This is why subdomain segmentation matters in the first place. A burned marketing.example.com does not have to drag down transactional.example.com if you set them up right. More on that in reputation inheritance between subdomains or IPs.

4. Clean the list before you send again

Remove anything that bounced in the last 90 days. Remove anyone who complained, ever. Remove role accounts (info@, sales@, postmaster@) unless they explicitly opted in. Run the list through validation to catch typos, dead domains, and traps that built up since your last cleaning.

This is not optional. If you resume sending to the same dirty list, you will be back in the red within a week.

What not to do

Do not try to send your way out. "Warming back up" by gradually increasing volume to the same bad list just stretches the damage over more weeks. Do not start using a new domain to dodge the problem unless you have fixed the list and process first. The new domain will burn the same way, only faster because it has no history.

Reputation repair takes 4 to 8 weeks of clean sending once the bleeding stops. If you skip triage, you are repairing nothing.