Feedback loop (FBL): how ISPs report spam complaints to senders

Feedback loop (FBL) is a channel through which a mailbox provider (ISP) tells the sender that a recipient clicked "Mark as spam." The provider packages that signal into a complaint report and delivers it to a special address the sender registered when joining the FBL program. The sender receives the report and is expected to remove the complainant from future mailings.

How FBL works technically

Most FBL systems use the ARF (Abuse Reporting Format) specified in RFC 5965. Each complaint arrives as an email with three MIME parts:

1. A human-readable description of the complaint.
2. A machine-readable block with the report type, complainant address, and other metadata.
3. The original message (in full or headers only) that triggered the complaint.

Senders parse incoming ARF reports automatically, extract the recipient address, and remove it from the list. Skip that step, and complaints keep coming. The provider's trust in your sending infrastructure drops steadily, eventually resulting in deferrals or outright blocks.

FBL at the major providers

Yahoo/AOL. Both offer a classic FBL through the CFL (Complaint Feedback Loop) program. Registration goes through a web form and requires domain ownership verification. Once registered, Yahoo delivers an ARF report to your designated address for every complaint.

Outlook/Hotmail. Microsoft runs two separate programs: SNDS (Smart Network Data Services) and JMRP (Junk Mail Reporting Program). JMRP is the FBL equivalent: register your IP ranges and Microsoft sends per-complaint reports. SNDS gives aggregate statistics without individual addresses.

Gmail. Google does not offer a traditional FBL with individual complainant addresses. Instead it uses the Feedback-ID header and surfaces aggregated complaint data in Google Postmaster Tools. You can see your overall complaint rate, but not who clicked "Spam" — Google shields that to protect user privacy.

Mail.ru. FBL is available through Postmaster Mail.ru. After domain verification you get access to complaint statistics. The report format may differ from standard ARF.

Why set up FBL

Without FBL you have no signal that a subscriber complained. They keep receiving your mail, keep complaining, and your complaint rate climbs. Once the rate rises across the board, messages start going to spam for everyone, not just the person who originally objected.

FBL cuts that loop short. One complaint, one removal, no repeat. Google's 2024 sender requirements set a complaint rate threshold of 0.1% — stay below that and deliverability stays stable.

Reducing complaints through FBL data

• Use double opt-in. Confirmed subscribers complain at a significantly lower rate than those acquired through single opt-in.
• Make the unsubscribe link easy to find. When recipients cannot locate it, clicking "Spam" is the next best option.
• Do not suddenly increase send frequency. An unexpected spike in volume irritates subscribers who were comfortable with the previous cadence.
• Segment by engagement. A subscriber who has not opened anything in six months is a good candidate to remove before they complain.

FBL and ESPs

If you send through an email service provider (SendGrid, Mailgun, Amazon SES), the FBL setup is usually handled on the ESP's side. The platform processes complaints automatically and marks affected addresses as "unsubscribed" or "complained." You do not need to parse ARF reports yourself. That said, it is worth confirming that your ESP is registered with all major ISPs and that complaints are actually being suppressed, not silently dropped.