BLOG

Feedback loops

There are a lot of different perspectives on Feedback Loops (FBLs) and “this is spam” buttons across the email industry.
Some people think FBLs are the best thing since sliced bread and can’t figure out why more ISPs don’t offer them. These people use use the data to clean addresses off their lists, lower complaints and send better mail. They use the complaints as a data source to help them send mail their recipients want. Too many recipients opted out on a particular offer? Clearly there is a problem with the offer or the segmentation or something.
Other people, though, think the existence of “this is spam” buttons and FBLs is horrible.  They call people who click “this is spam” terrorists or anti-commerce-net-nazis. They want to be able to dispute every click of the button. They think that too many ISPs offer this is spam buttons and too many ESPs and network providers pay way to much attention to complaints. The argue ISPs should remove these buttons and stop paying attention to what recipients think.
Sadly, I’m not actually making up the terminology in the last paragraph. There really are who think that the problem isn’t with the mail that they’re sending but that the recipients can actually express an opinion about it and the ISPs listen to those opinions. “Terrorists” and “Nazis” are the least of the things they have called people who complain about their mail.
One of the senior engineers at Cloudmark recently posted an article talking about FBLs and “this is spam” buttons. I think it’s a useful article to read as it explains what value FBLs play in helping spam filters become more accurate.

We have found over time that the most effective systems are those that learn to classify undesirable content based on feedback from users. The user is truly the best judge of what is and isn’t spam. The faster that consistent feedback becomes available, the sooner a filter can be re-trained to detect and respond to new attacks.

Spam filters are not there to stymie senders, their real job is to protect recipients. Faster adapting filters may seem like a bad thing, but they adapt in both directions. Send good mail that people want and it gets through. Send bad mail that people complain about and filters quickly adapt to block it.

2 comments

  1. dt says

    Is ok to use FBL as a statistic indicator but sometimes, individual complains can be wrong. It happen to me twice: client order online, spoke with him on the phone, send him a mail with some papers and surprise, he mark my mail as spam! One said he mark the mail from mistake, the other doesn’t answer but both made the payment and complete the order.

  2. The Proverbial Barry says

    my friend’s cousin’s brother who lives in new york got bit on the ass by a snake that crawled up from the sewer into the toilet but that doesn’t mean the sewers are really full of snakes.

Comment:

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.