Feedback Loops (FBLs) are also known as complaints, spam complaints, or abuse complaints.
What are Feedback Loops?
Feedback Loops (FBLs) are the systems by which a mailbox provider returns data about user complaints. This is usually based on actions of a user such as clicking spam or junk in the mail client. Mailbox providers offer this resource to help reduce unwanted mail from being sent to their users. Not all mailbox providers support feedback loops, and some have strict acceptance criteria to enrollment in their offerings.
The Feedback Loop Process
As a SocketLabs customer, you send your mail and it travels to the recipient's inbox. If the recipient clicks the spam button, that message is processed by the spam filter at the mailbox provider. Then, the mailbox provider generates a Feedback Loop and sends the FBL as a message back to SocketLabs in an Abuse Reporting Format (ARF).
Messages that are immediately placed in the spam folder by a mailbox provider can not be marked as spam by the end user and will not result in a Feedback Loop being generated.
What does SocketLabs do with Feedback Loops?
SocketLabs parses all received Feedback Loops and provides customers with associated data in a variety of means. The data for IP and Domain based Feedback Loops is provided directly back to customers in the Complaints report in the Control Panel. This data is also provided via Notification API (webhook) Events. Finally, SocketLabs offers to forward the original ARF Feedback Loops to customers for self-parsing and management. Aggregate Feedback Loop data is not provided directly back to customers at this time, but is used in StreamScore calculations.
SocketLabs also uses all Feedback Loops as a key part of StreamScore. Feedback Loops are one of the components for the Engagement sub-score within StreamScore. A high rate of complaints will have a significant negative impact on the Engagement score. For more details about the acceptable level of complaints see our knowledgeable article specific to that topic here.
You can monitor the Aggregate Feedback Loop offered by Gmail by DKIM signing your mail and registering your domains with Google Postmaster Tools.
What are the Different Feedback Loop Types?
IP-based Feedback Loops
The most common Feedback Loop Type is IP-based. Senders are able to register their IP addresses with mailbox providers and receive reports in the industry standard Abuse Reporting Format (ARF), which provides the sender with the full message, including the original recipient address. Feedback loops that are sent in this format allow for the identification and removal of subscribers who submitted the complaints from sender database.
You can read more about the concept for IP-based feedback loops as they are outlined in RFC 6449.
Providers that send feedback loops often only return a sample of complaint events to prevent abusive senders from removing all complainers from their lists.
Domain-based Feedback Loops
Much like IP based FBLs, Domain-based FBLs use the standard Abuse Reporting Format (ARF), which provides the sender with the original message in full. What is different is that the Domain-based FBL registration process is per domain, not per IP address. Senders must sign their mail with DKIM (DomainKeys Identified Mail) to be considered for inclusion in Domain-based FBL programs.
There are a few benefits to domain-based FBLs. First, senders using shared IP addresses are able register and receive feedback about their specific domain. Additionally, since DKIM authentication can survive message forwarding, mail that is forwarded then marked as spam can still generate a proper FBL to the original sender. For IP-based FBLs, the system forwarding the mail would receive the FBL and not the original sender.
Aggregate Feedback Loops
An Aggregate FBL is an alternative to IP-based or Domain-based FBLs. Aggregate FBLs roll up aggregate counts of complaints, often by IP, domain, or an identifier determined by the sender.
Providers of an Aggregate FBL format believe that spam complaints reveal too much personal information of their users. As a result, this format does not provide the sender with the full message and does not identify the user who complained about the email. Although recipient specific data is not provided, the aggregate statistics can provide senders with valuable insights about how their email program is performing.
What mailbox providers offer a Feedback Loop System?
SocketLabs has pre-registered all owned and operated IP addresses with all offered IP-based Feedback Loop services. SocketLabs has also pre-registered its own domain (email-od.com) and DKIM keys with all offered Domain-based Feedback Loop services.
Enterprise tier customers that do not have their messages dual-signed with a SocketLabs (email-od.com) DKIM signature can contact their account manager for more information regarding the Feedback Loops for which their domains have been registered.
- Google (Gmail only)
- Yahoo (Yahoo, AOL, Verizon)
- Italia Online (Libero and Virgilio)
- La Poste
- Liberty Global (Chello, UPC, Unity Media)
- Microsoft Consumer (Outlook.com/Hotmail/MSN/Live only)
- SilverSky (formerly USA.net/BAE Systems)
- Telecom Italia
- United Online
- Virgin Media
The following mailbox providers previously offered a Feedback Loop, but have since deprecated their offering:
- Roadrunner/Time Warner Cable