Security Metrics are Key Indicators of Security Health

Jody Brazil

Healthcare has many parallels to Information Security in that it attempts to diagnose and prevent negative outcomes.  However, if success is measured by “death prevention”, it seems to always fail eventually.  And if we take that same approach to security and only measure success or failure by “prevention of data loss”, we may have a lot of good days, but we will eventually fail, and perhaps catastrophically so.

Instead of waiting for these ultimate outcomes, there are many indicators that we can measure in both healthcare and IT security to positively affect our current situation.

For example, certain risk factors can be evaluated and measured related to prevention of heart disease (WebMD).  Some of these factors are outside of our control: sex, age and family history.  Other risk factors are behavioral and can have a significant impact on health outcomes including: diet, level of physical activity, and whether or not someone smokes.  Similarly in IT security, we cannot change the industry of our organization or the number of employees on staff; we also have very little control over bad actors who may choose to assail us.  But we do have control over what security technologies are deployed, employee education efforts and management of existing controls.  Measuring these items, particularly those over which we can control, will provide valuable information on how we can improve.

But it takes more than measurement to improve.  Knowing that one lives a sedentary lifestyle at the bottom of the activity scale will only improve their health outcome if they chose to change their behavior.  Metrics are merely an information source, but if they are used to inform and change behavior, they will affect outcomes.  Failure to measure these risk factors only perpetuates ignorance and leaves the results in the hands of hope and luck.

So what should you measure to improve your security health?  The answer depends a great deal on the specific situation of the organization.  There is little scientific evidence proving links between specific behaviors in IT Security and the potential outcome of data loss, at least not as conclusively as there are for health factors such as smoking and heart disease.  But there are many good sources of best practices including those provided by groups such as SANS, NIST, and ISO.  There is also valuable research containing correlations between certain activities and breaches in the Verizon Data Breach Incident Report among others.  The key is to understand your organization’s risk factors and then begin to measure them accordingly.

In most cases, some decisions about which risk factors should be address have already been made.  If not scientifically, at least these decisions were often made with a common sense approach.  The decision to deploy any defensive technology such as, firewalls, anti-virus, web filtering, intrusion prevention or others, were made based on the risk factors pertinent to your organization.  In most cases, these decisions were made to address the risks you can’t control (threats) in an attempt to reduce the exposure to those threats.

The next step is to measure the risk factors you can control, in particular, the management of the technologies you have already deployed.  As Gartner suggests, “more than 95% of firewall breaches will be caused by firewall misconfigurations”, clearly indicating it will take more than simply deploying a firewall to improve security, you must effectively manage the technology as well.  Measuring the effectiveness of ongoing management activity with key metrics such as complexity, permissiveness, unused rules, and other best practice measures is a great start.  To continue the healthcare analogy, this could be compared to exercise and heart disease: it may not be easy to get off the couch, but the risk in not doing it is severe and the benefits are proven.

Ultimately, it is the combination of all the risk factors that will determine outcomes in both examples: external, internal, those we can control and those that we can’t.  Measuring these indicators alone won’t protect you.  Measuring these risk factors and choosing to get off the proverbial couch and change IT security management behaviors will have a dramatic affect on the health of your security.