Today, we are sharing the relaunch of the free Abusix email feedback loop service, rebuilt from the ground up. Senders can now opt in with a single DKIM-FBL DNS record, alongside the RFC 9477 header, and we now send spam trap reports and reports on mail sent to our Blackhole MX service, which covers more than 350,000 domains. Within the first day of it being live, the number of reports sent more than tripled. Our view on this is simple: complaint data only reduces network abuse when it reaches the people who can act on it.
Complaint data has been stuck for too long
You’ve probably done this yourself. You open your inbox, find a newsletter you never signed up for, and click “This is spam.” Next week, the same newsletter arrives. And again the week after that. That happens when a complaint has nowhere to go, but a feedback loop fixes it. When someone reports a message as spam, the mailbox provider sends a report back to the mail service provider that sent it. The service provider removes that person from the list, and the mail stops.
For years, getting those reports was harder than it should have been. One centralized program handled feedback loops for most of the major mailbox providers. Senders had to register, hand over a lot of information about themselves, and get approved. The bar to entry was high, but it worked, so nobody built an alternative.
Then that program was productized, and the service was placed behind a paywall. The email community responded with two open ways for receivers to find out where to send reports automatically:
- RFC 9477, the Complaint Feedback Loop (CFBL) header. Senders add a header to every message saying where complaints should go.
- DKIM-FBL, a draft specification. Senders publish a DNS record on the domain they already use to sign mail with DKIM.
The standards solved discovery, but they didn’t solve the work. Every receiver still has to build a system to generate, format, and send reports, and every sender still has to set themselves up to receive them — many on both sides never really got there.
Reporting is the whole point
Some services in our industry ask you to send them your spam reports. They collect them, rank contributors on leaderboards, and use the data to improve their own blocklists and threat intelligence. That data has value — and it helps improve the quality of the blocklists, which subsequently helps block malicious traffic. We agree with that mission. But once it goes in, it goes nowhere else. The sender never hears about the complaint, and the network operator never gets a network abuse report. That’s where we differ.
“Every report we send gives someone the chance to fix a problem, whether that’s a sender cleaning up a list or a network operator stopping a spam run. That’s why we chose to send the data back out. The more receivers share, and the more senders opt in, the better it works for everyone on the internet.” — Steve Freegard
We think about it differently. Our mission is to be relentless in making the internet safer for everyone. Our vision is a global platform where every internet stakeholder can share data and make it actionable. A complaint that sits in a database isn’t actionable, but a complaint that reaches the sender, and a spam trap hit that reaches the network operator, is.
We also think not every complaint is the same, and reports should reflect that:
- User complaints are signals, not verdicts. People click “This is spam” on mail they simply don’t want anymore. Treating every click as network abuse would be unfair to legitimate senders. Therefore, user complaints go to the sender as feedback loop reports.
- Spam trap hits are evidence. Nobody signs a spam trap up to a mailing list. If a message lands in one, it’s there for a reason. Therefore, spam trap hits now generate both a full network abuse report to the network operator and a feedback loop report to the sender. Those are often different organizations, so both now get the chance to act.
The data shared with us through Data Channels also strengthens the evidence behind our own products. We’re open about that. The difference is that it doesn’t stop with us.
What’s new in Abusix Feedback Loop
We’ve run this free service through our Data Channels since early 2025. Running it taught us what was holding it back: senders had to be manually approved, we only supported the header method, and we used only part of the data we had.
So we decided to rebuild it from scratch.
| Report type | What triggers it | Who receives it |
| This is Spam | A user at a participating receiver reports a message as spam | The sender, as a feedback loop report |
| Spam Trap | A message hits our spam trap or a spam trap shared with us through Data Channels | The sender (feedback loop report) and the network operator (network abuse report) |
| Blackhole | A message is sent to one of the 350,000+ domains on our Blackhole MX service | The sender, as a feedback loop report |
Each report type has its own clearly labeled email template, so recipients know exactly what they’re looking at, with the exception of spam trap reports: They go through extra redaction, because a spam trap that’s been revealed is a spam trap that no longer works.
Three further changes make it easier to take part:
- No manual approval. If you’ve opted in using either standard, you get reports automatically.
- DKIM-FBL support. RFC 9477 remains fully supported. You can now also opt in with one DNS record on your DKIM signing domain, with no changes to your sending platform.
- Your choice of format. Receive reports as ARF/MARF or X-ARF, specified in your header or DNS record.
“Adding a header to a mail platform turned out to be far harder for many senders than expected. DKIM-FBL is a DNS change. It’s a five-minute job, so if you want feedback loop reports, it’s now been made so much simpler to do.” — Steve Freegard
The result: we’re reporting double the amount of traffic compared to two days ago. Including Blackhole traffic, reporting volume has more than tripled. The largest share of new reports we have observed goes to Amazon SES, Gmail, Klaviyo, ConstantContact, SMTP.com, Mailkit, SocketLabs, and AWeber.
How to set up your own email feedback loop
If you send email, whether directly or through an email service provider:
- Check whether you’re already opted in. Some large email service providers already opt in on their own signing domains, so you may already be receiving reports through them.
- Opt in on your own domains. Publish a DKIM-FBL record on your DKIM signing domain, or add an RFC 9477 header to your messages.
- Choose your format. Specify ARF/MARF or X-ARF in your record or header.
- Be ready to process reports. Each report includes the original message headers. Use the Message-ID to match a complaint to the recipient and campaign, unsubscribe that person, and track complaint trends in your dashboard.
- Treat Spam Trap and Blackhole reports as list hygiene alerts. They usually point to addresses you never verified or a collection process that needs review.
If you receive email as a mailbox provider, ISP, or hosting provider:
- Set up a Data Channel. It’s free, and you don’t need to build your own reporting system.
- Tell us what you’re sending. Mark each channel as user feedback or spam trap data, so we report it the right way.
- Forward your data. We generate, redact, and send the reports to every sender and network operator who has opted in.
Every report that reaches someone who can act is a step forward
The service is live now, and reports are already flowing. If you send mail through a large email service provider, you may already be seeing more of them.
Feedback loops have always depended on two things: receivers willing to share data, and senders able to receive it. We’ve removed the cost and most of the effort on both sides. Opting in with DKIM-FBL takes five minutes. Sharing data takes one Data Channel. Both make the internet a little safer for everyone.
Get started. Learn how to set up your Abusix feedback loop in five minutes here.