How Spam Email Filtering Techniques Are Being Applied to Enterprise Voice

Email spam used to be a huge problem. It's under control today, thanks to spam email filtering, but a new productivity and security challenge is at hand: robocalls and voice spam.

It’s funny how history can sometimes repeat itself, but in a slightly different way. Back in the late 90s, when the Internet was really starting to catch on, one of the single biggest annoyances that people suddenly found themselves having to deal with was spam. I know people who actually abandoned their email accounts and got new ones simply because the spam in their inbox had become so overwhelming.

Twenty years later, spam hasn’t gone away. Even so, it is far more manageable than it once was, thanks to better spam email filtering technology. On any given weekday, for example, I receive several hundred spam messages. Even so, fewer than a dozen slip through and make it into my inbox. Nobody is claiming that spam email filtering is perfect, or that spam has been eradicated, but the technology is far more effective than it once was.

Of course, nature abhors a vacuum: Enter robocalls and voice spam.

In fact, robocalls (and scam calls from live callers) have become so pervasive that many people, including my good friend and YouTuber Kenzie Claire, won’t even answer calls from numbers that they do not recognize. As annoying as these calls can be, they aren’t just a personal inconvenience. Robocalls and scam calls can present a huge problem in the enterprise. In addition to the costs associated with tying up phone lines, there are costs in lost productivity. When employees receive an unwanted call, they waste time not just answering the call, but also getting back to what they were doing before they were interrupted.

An article from PayScale.com breaks down the per-minute pay for employees in various positions. The median amount is about 45 cents per minute. With that in mind, imagine an enterprise-class organization with 5,000 employees, each of whom loses a mere 3 minutes per day to robocalls. At 45 cents per minute, that works out to $6,750 dollars of lost productivity every single day. When you consider that employees generally work about 250 days per year, this figure translates into about $1.7 million annually in lost revenue. Clearly, it is in the enterprise’s best interest to put a stop to the disruptive robocalls.

Of course, it isn’t just the robocalls that are problematic. Scam calls are also an issue. Who among us hasn't received a call from a scammer posing as some government agency or other? I couldn’t even tell you how many times I have received calls from scammers who tell me that unless I use iTunes gift cards to pay overdue taxes, I am going to jail.

These types of calls are annoying, for sure, but enterprise organizations often have to deal with a bigger problem.

Many large companies lease dedicated telephone exchanges. For instance, one of the companies that I used to work for owned every phone number starting with 580 in a certain area code. Because the scammers knew this, employees would periodically receive targeted scam calls from someone pretending to be in HR, IT or finance.

The good news is that the telecommunications companies are finally beginning to realize just how harmful robocalls and scam calls can be, and have begun creating filters similar to spam email filtering technology.

A company called Mutare, for example, has designed an innovative filtering system designed to work in conjunction with the company’s voice mail systems. When an employee receives a call from the outside world, a screening proxy server analyzes the call in an effort to determine whether the call should be allowed. The caller ID information is compared first against an enterprise managed whitelist, and then against an enterprise managed blacklist. If the caller ID information does not appear on either list, then it is compared to a cloud database of phone numbers that are known to be associated with undesirable calls. If the call passes this check, then it is forwarded to the employee’s phone. Otherwise, the call recipient’s phone never even rings. Robocalls are simply dropped.

Although Mutare’s approach will likely help its customers receive fewer unwanted calls, I tend to think of it as a first-generation solution. The problem with Mutare’s approach is that robocalls and scam calls often use spoofed phone numbers. This spoofing could lead to robocalls getting through the filter, while legitimate callers may be blocked simply because they are unlucky enough to have had a robocaller spoof their phone numbers at some point in the past.

It telecommunications providers are to truly win the war against these unwanted calls, they will need to look once again toward spam email filtering solutions. Spam and phishing messages often use spoofed sender information, and yet the spam filter is able to identify the message as spam anyway. It accomplishes this by looking at other message characteristics and then using AI or heuristics to determine the probability of the message being spam. Companies seeking to put a stop to robocalls could use the same approach to more accurately identify voice calls.