|By Mark Bregman||
|March 17, 2014 09:45 AM EDT||
Given the mountains of data now floating around, it is perhaps inevitable that the very function of data analytics is seen as somehow intrusive. There's a constant glut of reports, columns and other stories bemoaning the lack of data privacy - and at times, they're entirely justified. Organizations have a solemn duty to protect their customers' data, and those that don't implement the proper safeguards deserve to be vilified.
But beneath the surface lurks another dimension of this discussion that is often overlooked. Ethical and effective data analytics enhances security. Ethical and effective data analytics protects not only the institutions that possess the data, but also the consumers that data reflects. Ethical and effective data analytics serves a good purpose.
Let's be clear about the parameters of this argument. Data doesn't exist in a vacuum - it's generated on an ongoing basis through multiple activities, created in numerous formats and comes in through a variety of channels. At any given time, it is being analyzed and used (and occasionally misused) to serve many different needs.
Of course, when done right, information services and analytics represent a key driver of most business decisions. Actionable intelligence based on real data doesn't just augment gut instinct; it leads to quantitative thinking that supports strategic initiatives, enables tactical outreach and boosts the bottom line. Perhaps most important, it enhances information security so as to protect customer privacy and prevent operational and brand damage.
High-profile assaults on retailers like Target and Neiman Marcus, or clandestine downloads of classified information from the National Security Administration (NSA), make more news than inside-the-infrastructure DDoS attacks, but the latter is even more insidious. There are over 2,000 DDoS attacks every day. Some 65 percent of all organizations see three or more attacks each year. While the devastation is certainly felt on an organizational level, the financial impact is just as significant: DDoS attacks can cost up to $100K an hour.
DDoS mitigation can be an enormous challenge. Making an accurate distinction between normal, benign Internet traffic and malicious activity that could be the root cause of a potential DDoS attack is critical, but it's not easy. This is in part because DDoS attacks, especially when they serve as the front line of advanced targeted attacks, are remarkably sophisticated. They rely on stealth techniques that go unnoticed within the enterprise for long periods. They're highly customized, based specifically on each target's infrastructure and defenses, and can often defeat defense strategies that rely primarily on signature-based detection. Then of course there's the cloud. When attacks become super-sized, the defensive strategies in place must have the capacity to scrub far larger volumes of bad traffic.
This is why information services and analytics are so crucial. They can boost awareness and reaction time to particular situations. When it comes to leveraging Big Data within the enterprise to help identify breach attempts, it's still early days. According to a January 2014 report from Gartner, eight percent of enterprises are using data and analytics to identify security flaws. But there's reason for optimism - the same report also estimates that within the next two years, around 25 percent of enterprises will leverage Big Data for security purposes.
It is this same pattern-searching approach that the enterprise should take when it comes to DDoS mitigation. Proactive site monitoring on a continuous basis - in particular with a centralized view of traffic patterns - enables organizations to identify anomalies and threats, before they become real problems. For example, in the case of a custom application being exploited for a directed attack to steal customer data, the detection solution must be able to identify and highlight the fact that there's a new kind of application traffic on the network.
This might be a new concept to enforce at the enterprise level, but this is really something that banks have been doing for years with regard to fraud protection services. Banks monitor a person's transaction activity, and when a purchase is made that does not fit the usual spending behavior, it is stopped and flagged with the customer. The same thing should - and will - happen at the enterprise level.
It's easy to see why information services and analytics are too often seen as a potential invasion of privacy. Data privacy is vital, and it should rightfully be a corporate priority. However, in the ongoing effort to secure data, the right kind of analytics can be the best weapon of all.