Feedback Loop when the receiver tells you it's spam

The major FBL programmes

• Yahoo CFL (Complaint Feedback Loop). Now under Yahoo's post-Verizon stewardship. Open to senders meeting basic authentication and volume criteria.
• Microsoft JMRP (Junk Mail Reporting Program). Covers Outlook.com, Hotmail, Live, MSN. Enrolment via the SNDS (Smart Network Data Services) portal.
• Microsoft SNDS itself. Not strictly a complaint FBL, but the same portal exposes per-IP volume and complaint rates which is the next-best thing.
• Comcast SmartFBL. Covers Comcast email and Xfinity Connect customers. Smaller in volume now but still relevant for US consumer mail.
• Verizon (now Yahoo) and AOL. Consolidated under the same Yahoo programme post-acquisition.
• Gmail. Notably does NOT offer a per-message complaint FBL. The closest substitute is Gmail Postmaster Tools, which exposes aggregate spam rates without per-message granularity.
• Long tail. Zoho, GMX, Mail.ru, NetEase, La Poste, and other regional providers each run their own. Spamhaus's DBL feedback programme is sometimes lumped in here too, though it functions differently.

How an FBL report looks

Most FBLs deliver reports in ARF (Abuse Reporting Format, RFC 5965) — a multipart email containing the original message headers, the complainant's address (often obfuscated), the timestamp, and the type of complaint. Your platform parses these as they arrive (essentially listening on a dedicated mailbox for ARF-formatted inbound mail) and routes the action automatically:

• Suppress the complainant immediately (one-strike, no exceptions). They marked you as spam; never send again.
• Increment your complaint-rate counter for the affected sending IP and domain.
• Cross-reference with the original campaign or activity that produced the message.
• If complaint rate crosses a threshold (typically 0.1%-0.3%), trigger throttle or pause.

Why FBL participation isn't optional

Two reasons.

First, the practical one: if you don't process FBL reports, you keep sending to people who marked you as spam. They mark you as spam again. Their mailbox provider escalates. Now you're in the spam folder for everyone on that provider, not just the original complainer. The cost of ignoring FBLs compounds fast.

Second, the political one: mailbox providers treat FBL enrolment as a proxy for sender intent. A legitimate sender enrols and processes the reports. A spammer can't be bothered. When Yahoo or Microsoft evaluates your sending reputation, they note whether you're paying attention. Senders not enrolled in major FBLs are treated more suspiciously.

The DKIM tagging trick

FBL reports often obfuscate the recipient address (to protect the user). To still know which campaign or recipient generated the complaint, mature senders tag outbound DKIM signatures with a campaign identifier. When the report comes back, you parse the original DKIM header from the embedded message, extract the tag, and route the complaint to the right campaign-level counter even without the recipient address. This is how good platforms maintain per-campaign complaint rates despite the privacy hedge.

How sendmsg.io handles FBLs

Cortex is enrolled in the major FBL programmes on behalf of the platform. Incoming ARF reports are parsed in real time, the recipient is suppressed instantly, and the complaint feeds straight into the per-activity, per-domain, and per-mailbox-provider reputation temperatures. The DKIM campaign-tagging is automatic, so per-campaign complaint visibility holds even when the recipient address is obfuscated in the report. If a campaign's complaint rate crosses 0.1%, graduated throttle kicks in; at 0.3%, the campaign pauses pending review. You don't have to plumb anything yourself.