Facebook Likes and Personality Prediction: An Untapped Source for Further Hyperpersonalization?
The predictive power of Facebook knows you down to your personality. This is probably no surprise to us BIM students – the information density we share on Facebook surely provides enough data for Facebook to be able to predict who we are. However, what if your personality could be predicted simply through your Facebook ‘Likes’?
In a study published in PNAS in January 2015, Scientists from Cambridge University built a predictive online application to prove that computers can more accurately judge peoples personality based on their digital footprints (in this case Facebook Likes). Solely through analyzing ones Facebook Likes, the applications algorithm could immediately estimate their education and relationship status as well as present an intelligent guess about their gender, intelligence, politics, religion, life satisfaction and sexual orientation and preferences. Most interestingly, the application could also estimate ones ‘Big Five’ personality traits (openness, conscientiousness, extraversion, agreeableness and neuroticism).
The results: the algorithm of the application could more accurately predict the participants personality than a colleague through just 10 Likes; better than a friend or roommate through 70 Likes, better than a family member with 150 Likes, and a partner with 300 Likes.
The study emphasized that the results could be used by police to monitor potential criminals on social media – even before they commit a crime. However, I am a businesswoman, and though the findings are extremely valuable for societal welfare, I believe these findings on personality prediction could also have HUGE commercial potential. Countless studies have shown relationships between consumers personalities and their shopping and search behaviors. If firms are able to quickly complete estimated, yet accurate, personality descriptions of their clients through simple metrics such as Facebook Likes, this further understanding can be leveraged to predict their clients behavior and cater their unique needs.
With firms being able to collect and process data at an ever higher speed, the doors have opened up for firms not to only understand their customers on a demographic level- but also a personal one – known as hyperpersonalization. Today firms with the correct data collecting infrastructures, are able to offer their clients personalized and targeted products, services and content. This usually involve delving into rich datasets about their consumers attributes and behavior. Firms can then use this data to reach out to the right buyers at the right moment with the right product. Indeed, hyperpersonalization is highly profitable. A research report published by VentureBeat showed the impressive results of reaching out to customers as unique individuals. For example, simply by offering personalized web content, one firm increased page views on their site by 300% and another firm saw their conversion rates boost by 219%.
Inspired by this new insight about the ease at which algorithms can predict personality, I believe it may not be long before firms will also be able to predict their customers personality – allowing further hyperpersonalization. Indeed, so many brands have access to our Facebook accounts today, this is easy data to collect. Moreover, this understanding may not only be able to allow firms to further predict customer preferences, but also how and where they may want their preferences to be presented to them. For example, a study by Chen and Lee (2008) found that consumers with higher levels of agreeableness and conscientiousness were more focused with fulfilling utilitarian needs when online shopping whereas consumers with higher levels of emotional stability, openness, and extroversion, were driven more by the experiential and hedonistic values. In marketing, this could be practically translated to separate marketing campaigns to different personality types – one campaign emphasizing their products utilitarian aspects and another the hedonic features. The possibilities are limitless!
It will not surprise me if big firms are already onto this and offering us personalized unique product, service and content propositions based on our personalities right beneath our noses. I have researched this online and have not been able to find any articles saying that analytics teams are also taking personality into account. Maybe some of you know?
Interested in finding out your predicted personality? You can take the predictive test and find our through Apply Magic Source test (the same used in the Cambridge study) to find out your predicted personality based on your Facebook likes here.
Youyou, W., Kosinski, M., & Stillwell, D. (2015). Computer-based personality judgments are more accurate than those made by humans. Proceedings of the National Academy of Sciences, 112(4), 1036-1040.
Chen, S. H., & Lee, K. P. (2008). The role of personality traits and perceived values in persuasion: An elaboration likelihood model perspective on online shopping. Social Behavior and Personality: an international journal, 36(10), 1379-1399.
T.G. Stacey, SID: 441696ts