Monthly Archives: September 2010

Privacy is the center piece of a lot of jitters in the integrated web experience today. Whether its Google showing you contextual advertisements or Facebook Open Graph compliant sites giving you a tailored experience, you are putting your privacy on line. What’s the point? It’s to get personalization. Personalization has become a necessity in the world today. With content overflow in the world around us, the only way to stay on the top of your areas of interest and expertise is by getting a personalized experience.

There is an interesting trade-off between personalization and privacy. While on one hand privacy is important to protect your identity from the increasingly vulnerable web, personalization can make you more efficient and presentable than anything else. So the important question here is, where to draw the line? In order to define that, it is critical to understand how much personal information is needed to provide a personalized experience. There are degrees of personalization and quite expectedly, different degrees of personalization requires different levels of information invading privacy at different levels.

One of the most common forms of personalization (beyond contextual advertising) is behavioral targeting where browsing behavior of an individual is used to personalize the experience online. This is the most simple form of behavioral targeting, where the advertisers and content providers do not use any personally identifiable  information about an individual and provides contextual relevance just based on the browsing history. In this case, the privacy is marginally compromised and the degree of personalization you get can make your browsing experience very optimal.

The next step in providing deeper personalization is to integrate an individual’s profile information with the browsing behavior. You experience this kind of profile based behavioral targeting at online stores like Amazon. Amazon knows a lot about the customer’s profile which includes their shopping pattern, their wish list, their preferences and much more. Combining this information with the browsing behavior of a customer, both on the online store and elsewhere on the world wide web helps the online store provide you with personalized shopping recommendations. Another arena where this type of personalization is experienced is at social networking sites like Facebook. Facebook collects a lot of information about the customers through their profiles, their status updates, their friends, their groups and so on. It uses this information combined with the behavioral traits of the customers to provide personalized advertising and promotions. Now let’s throw privacy in the discussion. In case of personalization at these domains, the content providers are using the information the customer has explicitly provided them. In other words, the customers can control their privacy invasion by controlling the amount of information they share with these domains. The trade-off is in terms of personalization, but as far as privacy is concerned, it’s completely under control of the customer and the trade-off is very explicit.

Move it a step further and we come to personalization on the network using an individual’s behavior and their profile information available with the network creator. Think of it as expansion of the previous scenario where the single domain has now diluted to contain several websites and online properties that agree to share the profile information of the customer to provide richer personalized experience. At one end this network based targeting provides a great personalized experience to the customers but on the other hand has raised the privacy concerns more than in any other case. In the defense of the networks, they are doing nothing without taking customers’ consent, but more often than not the customers are clueless about the consent they are providing the content providers. So how can privacy issues be addressed here? I believe the best way to do this is by educating the customer about what they are doing and by explaining them what they are gaining and what’s the cost. By doing that, it will be up to every individual to decide what cost in terms of privacy they are willing to pay for gains in terms of personalization!