The Evolution Of Spam Filtering And Deliverability

Posted on Oct 3 2017 - 4:19pm by Editorial Staff

Email communication has been an important part of the marketing toolbox for the past quarter of a century, and is as much so today as it was in the 1990s. Like every form of direct communication, whether it is physical mail, cold calling or the good old fashioned foot in the door, part of the challenge is targeting your efforts at those who are interested in what you have to say, and convincing them that you are worth listening to.

When it comes to email, this brings us straight to the concept of spam. It has been around for as long as the internet, and is as relevant today as it has ever been. Spam filters are great for keeping our inboxes clear of junk and potentially harmful links and malware, but they can provide a real problem for genuine marketers who are simply trying to reach out to their customer base.

Like every aspect of software, spam filtering has constantly evolved to meet new demands and the latest technology. Today, the focus is very much on cloud-based solutions, and tools such as the everycloudtech cloud spam filtering service have grown up around the new technology. But all of these are built on foundations that go back to those earliest years. Let’s take a look.

The pioneering years

Let’s face it, back in the 1990s, it was a free for all. We were all leaping aboard the internet, and we neither knew nor cared where it was taking us. Hotmail was born in 1996, and became the must-have. Genuine marketers jostled with the scammers and spammers to get our attention, and all we could do was try to avoid the elephant traps as best we could.

The first filters

Internet service providers soon saw that everyone’s email was starting to fill up with junk, viruses, porn and all sorts of other unwanted detritus, so they set about doing something about it. This was effectively to create the first blacklists, blocking senders who were known to be trouble, and to filter out messages as best they could on the basis of likely keywords. This was effective to a certain extent, but led to many genuine mails being mistaken as spam.

The arrival of the postmaster

ISPs tried to improve things by introducing whitelists, so that they were not just looking to filter out the bad stuff, they were actively pushing through emails that were definitely OK. Again, this represented a positive step, but the crude systems were easy for spammers to circumvent.

Lists and reputations

By the late 00s, we had entered the Gmail era, and the concept of reputation became important. The way it works is that the more users flag a sender as a spammer, the lower his reputation. Similarly, those senders who actively manage their distribution lists are going to improve their reputations.

Engagement metrics

This brings us to the present day, and the world of engagement metrics. ISPs track what happens to a message, and use this to understand whether it is wanted. For example, it might be deleted unopened, ignored unopened, opened and read, opened and the content clicked on or opened and replied to.

Each of these provides useful information for more accurate spam filtering, and the current advances in big data analytics and machine learning will make the next generation of spam filters even more accurate and valuable, for senders and receivers alike.

About the Author

Editorial Staff at I2Mag is a team of subject experts.