We hope that this will be a forum for conversations about the incredible opportunities and challenges of Social Psychographic Targeting for advertisers and publishers on the Social Web.   We welcome your feedback and comments – after all, we believe so much in the power of the social features of the Internet such as blogging, social networks and interactive discussion forums just like this one that we have built our company around better understanding the data presented in these arenas.  One of the great things about User-Generated Content is that it allows people with good ideas to share them with the world easily, and we at Peerset know that no one person or company has a monopoly on good ideas.  We love hearing from anyone and everyone because you never know where that nugget of wisdom may lie.  Yes, that means YOU.

Since this is the first posting, let's start with a little background. Before Peerset's current incarnation, and prior to the explosion of social networks like MySpace, Facebook and LinkedIn, the founders of Peerset had been researching and developing different ways to use computers to help stimulate creativity and explore new ideas.  We took as our starting point the reigning idea that the core feature of creative thinking is the ability to see connections where they might not have been noticed before. Many studies have shown that truly creative people do not come up with their best ideas out of thin air, but rather they put existing concepts together in new and unusual ways that lead to positive innovations.  With academic training in such fields as linguistics, philosophy, physics and cognitive science, we naturally took a top-down approach to the problem of creativity and built a massive Ontology called the Concept Specific Ontology (CSO) that allowed ideas to be interrelated along many different axes in order to highlight the connections between concepts that are not normally evident.  The rise of social media occurred while we were building this system and stimulated our own burst of creative insight: interesting connections are being made by millions of people everyday on social networks through UGC found in profiles, tags, blogs, instant messages, wall posts etc…  We realized that while it is difficult for a human observer to make sense of all the connections being made, the power of computer processing could be harnessed to show patterns of connections that could be used to recommend new pathways in almost any domain.

We soon realized that by looking at constellations of ideas that were found more commonly amongst users, we could get a good picture of the kind of psychological makeup with which various segments of society are endowed. In other words, whereas looking at ideas, interests, hobbies and the like that occur rarely in certain contexts could suggest what we might call "creative" or "innovative" ways of combining these ideas, ones that occur together commonly can be used to get a picture of how people belonging to various segments tend to think. We thus began work on a system to supplement the CSO that we called the Interest Correlation Analyzer (ICA) and out of this, Peerset's Social Psychographic Targeting was born.

Psychographics is the study of what kind of personality, opinions, interests and attitudes are held by different segments of the population. The theory is that if something is known about a person's interests or demographic status, then, based upon analysis of others like them, recommendations can be made that are likely to attract and engage them. It is quite fascinating to see this operating in action, because we are brought up in this society to believe that we are all individuals, which of course we are in many ways. But on the other hand, people do tend to cluster into psychographic groups of like-minded people and if some in this group respond favorably to a marketing campaign, the others are likely to do so as well. We must emphasize the word "likely" because this is all a statistical play. It is of course perfectly possible that an individual who fits well into one segment does not in fact like some of the things that the other people in this segment enjoy. But there is a much greater possibility that they will like it than not, and in an imperfectly understood world, that is the best we can hope for.

What seems pretty clear, though, is that social networks and social features of the Internet in general offer an unprecedented opportunity for both building psychographic profiles of various segments, and for marketing to them. Unlike more traditional methods of building psychographic profiles, such as conducting surveys on a few thousand people and then extrapolating from this, social media allow us to build the profiles using data from the actual people who are then going to become the audience that receives the targeted advertisements or other content. The huge amount of users of these sites also allows for the collection of sample sets that provide unprecedented accuracy because of their size and breadth. Furthermore, the quality of the data is not affected by any experimentally produced biases because the data that people offer in their social profiles is for the most part unprompted and emerges organically at the time and in the form of the user's choosing. This freedom, which is one of the great strengths of social media, also leads to one of the weaknesses from the advertiser's perspective, which is the unstructured and "messy" nature of much of the UGC. A fair degree of linguistic intelligence is required of any ad serving technology that would attempt to digest this data and make sense of it. This may be one of the reasons that the promise of Social Advertising has not yet been fulfilled. However, if care is taken to extract, process and utilize data in the right way, there can be little doubt that social media offer a unique and unprecedented channel for engaging with interested customers in ways that lead both to successful monetization and high-quality user experience.