The following is from Peter Rothman (@dangrsmind), one of my regular contributors. It contains a controversial and interesting perspective for the web analytics community. We invite your comments on the matter.
In a recent blog posting at the WAA Thought Leaders blog, Eric Peterson of Web Analytics Demystified proposed a ten-point code of ethics for the web analytics community (http://waablog.webanalyticsassociation.com/2010/09/web-analytics-code-of-ethics.html) Many in the analytics community will be supporting this idea, but the reality behind the scenes may not turn out to be so clean.
Unfortunately this effort is destined to fail. It should be recognized that producing an industry code of ethics is an important and laudable idea. Employing a code of ethics is not just about being a moral person (although that is also a worthwhile goal in itself): it is about making effective decisions and, in particular, effective business decisions. Recognizing this important point is the beginning of a more robust approach to this issue of ethics as well as the path to making better business decisions.
First, recognize that the problem is much worse than you think. News reports of unethical practices in the web analytics community are just one story that is being told about business misconduct and poor or non-existent ethical practices in business. Consider Enron, the global financial meltdown, the sub-prime loan crisis, Bernie Madoff, etc. etc. As a result of this pervasive environment of mistrust, the average person is pre-disposed to believe the worst about any business sector or executive caught with their hands in the cookie jar.
Second, the problem doesn’t just start in the analytics or marketing group. The CEO is inevitably driving the company strategy and leading by example as are the board of directors and other C-level executives of the corporation. If the C-level leadership of a company is setting high ethical standards themselves and emphasizing ethics as a practice in their organization, the other members of the team are likely to follow suit. But again we see that the daily news is full of stories of executive misconduct and malfeasance, the biggest recent story being about the ousting of HP CEO Mark Hurd, but there are many less well known stories as well. While commendable, setting practices in the analytics community doesn’t change the behavior in the board room.
Third, while members of the white hat analytics community may adopt a code of ethics, such as is proposed by Peterson, there are others who will just laugh all the way to the bank. Recognize also that for users of analytics services, the vendor is a risk in this area; e.g. in the so called “Zombie Cookie Case,” Disney, and MTV etc. are being sued because of their use of a third party service which they employed on their sites. Cookie regeneration from Flash was just one feature of these services that has come to light. This is the tip of the iceberg perhaps, but it is also not entirely clear what Disney et al knew about this or any specific feature of these products at a deep level. Vendors should consider what might happen if customers decide this entire category of service is too risky due to some high profile litigation or news story.
Fourth, these sorts of legal cases also point to a deeper weakness in the overall business of web analytics itself. Note that some of these products exploit security holes in browsers, security model weaknesses in Flash, etc. These same type of hacks which are also used by computer hackers and computer virus authors to gain unauthorized access to systems. As a result, there are many people who are interested in blocking exactly the sort of techniques that web analytics vendors sometimes require to track people. These include software providers of operating system software, major security software vendors, and the U.S. Federal Government. That’s a bad list of people to be working against you isn’t it? And further, this uncertainty about how long an important technical method will continue to work may drive some analytics companies to cut even more corners as the emphasis is increasingly towards the quick buck and early cash out. The entire situation is further raising the risk of a high profile data breach or cybercrime story originating in the web analytics space.
Finally, the reality is that consumers don’t want to be tracked and they generally don’t support the use of web analytics techniques when they learn about them. They don’t understand any benefits they might receive directly or indirectly from allowing this tracking of their online activities. Even in ethical companies that are working to protect consumer identities and information, there is a possibility of breach or unintended disclosure. Again, one big news story is all it takes to tip the sentiment even further negatively on the use of these technologies. To the extent that so many vendors still remain opaque about how they exactly do what they do, customers using these services face yet another unknown set of risks.
So while an industry code of ethics is a good idea, it has to remain just a starting point for cleaning up a broader set of issues. Other important things for the analytics community to consider are improved security and privacy policies and deploying related technologies to secure any private data captured, using “anonymization” of collected personal data wherever possible, deploying data access controls, and developing and enforcing internal data retention/deletion policies and making all of these transparent to both customers and consumers. Until both ethical standards and this broader set of technical and business related issues are seriously addressed the industry really does seem to be sitting on a ticking time bomb.
One last note: it is my belief that ethics starts at home with personal thought and behavior. Therefore developing an ethical code is–and must always be–a personal process. An industry organization can suggest some points of shared ethical agreement, but this is useful, however only each of us individually can develop and practice an ethical code. And at the end of the day one only has to face only oneself, and the consequences of our decisions and action in relation to that ethical code.
This book was given to me and I recommend that anyone with an interest in improving their personal practice in this area acquire a copy of Ethics for the Real World: Creating a Personal Code to Guide Decisions in Work and Life by Ronald Howard and Clinton Korver and use this book to create your own personal ethical code. This book is highly recommended and is a good place to start for anyone wishing to extend or further evaluate Mr. Peterson’s list of ten ethical points for the web analytics community.