Welcome to the SecAware blog

I spy with my beady eye ...

28 Nov 2012

SMotW #34: homogeneity

Security Metric of the Week #34: organizational and technical homogeneity

The degree of homogeneity (sameness) or heterogeneity (variation or variability) within the organization and its technologies affects its aggregated information security risks, in much the same way that monoculture and multiculture crops may face a differing risks from natural predators, parasites, adverse environmental conditions etc.  A particular mold that successfully attacks a certain cultivar of wheat, for example, may decimate a wheat field planted exclusively with that cultivar whereas it may not take hold, making little impact on a neighboring field planted with a mix of wheat cultivars differing in their susceptibility or resistance to the mold.  On the other hand, under ideal conditions, the monoculture crop may do exceptionally well (perhaps well enough to counteract the effects of the mold) where the mixed crop does averagely.  

Homogeneity of technologies, suppliers, contracts etc. increases an organization's exposure to  common threats - for example, serious security vulnerabilities in MS Windows may simultaneously impact the millions of organizations that rely on Microsoft's products.  On the other hand, homogeneity means standardization, lower complexity and ‘economies of scale’, generally generating substantial business benefits.  It is clearly in Microsoft's commercial interests to be seen to address serious security vulnerabilities in its products urgently, or risk mass defection of its customers (those who aren't entirely dependent, at least!).

The overall PRAGMATIC score for this candidate metric is mediocre:


The metric rates poorly on both Timeliness and Cost due to the difficulties of gathering and analyzing suitable data with any kind of precision.  However, a quick-and-dirty low-Accuracy assessment might be sufficient get this issue raised and discussed at the top table, which might actually be good enough (we're hinting at the measurement objective - an issue we have hardly mentioned in the blog but which is covered at length in the book).  The metric may perhaps be measured using scoring scales that we have discussed in several previous blog postings, for instance.

Sitting at 40%, the Actionability rating is also depressed for two distinct reasons: 
  1. It is not entirely clear what constitutes an 'ideal' amount of homogeneity, since, as we have just said, there are pros and cons to it;
  2. There are obvious practical constraints on management's ability to change the organization's homogeneity even if they wanted to do so.  Senior management might institute a supplier diversity policy, for instance, but there is likely to be considerable inertia due to the existing portfolio of suppliers currently contracted.  In many cases, there will be overriding commercial or technical reasons to retain the current suppliers, on top of the natural affinity that emerges through social interaction between individual employees and their supplier contacts.
Bottom line: this candidate metric is unlikely to make the grade for Acme Enterprises Inc., but it may be valuable elsewhere.

No comments:

Post a Comment