Microtargeting is a methodology linked to marketing that aims to influence the decisions of clients, consumers, or the general public.
It consists of the management of huge amounts of data, in which common patterns are sought through selection criteria that consider inclinations, interests, concerns, socio-economic situation, educational level, age group, etc., in order to create segmentations of the total set.


The people of each of these subsets will be the recipients of messages designed to suit them, with a high degree of personalization, thus achieving an increase in the impact and the expected response. The effectiveness of the technique is directly proportional to the precision achieved in the knowledge of the individual characteristics of the people to whom it is addressed.

Background


In 2008, Michal Kosinski and David Stillwell, then-doctoral students at the Center for Psychometry at the University of Cambridge, developed a small application for Facebook called MyPersonality, which from a short series of questions based on the categories of psychometry known as the model of the big five (extroversion, open-mindedness, responsibility, affability, and neuroticism) and the correlation of the answers with other generic personal data added to the "likes" in their participation in the social network, they managed to describe with a high degree of precision the personality profile of each of the respondents.


A later study carried out on a universe of more than 80,000 volunteers revealed that - applied the appropriate algorithms -, a computer system that analyzed only 10 " likes " could predict the personality of an individual with more accuracy than a coworker. As the number of analyzed " likes" increased, the accuracy of the computer system increased, even exceeding the accuracy of people in the inner circle.


At the end of 2016, a new set was unveiled that incorporated the values ​​and needs into the initial five categories. The model developed by IBM includes 47 dimensions, making it much more powerful in formulating predictions about the personal characteristics of the profile analyzed.

Data Management

At present, an increasing number of daily personal actions are likely to be registered in any of the multiple existing reservoirs.

Credit card transactions; free user account creation on web portals; applications, contact lists or emails, among other possibilities offered by smartphones; the GPS present in mobile devices; In addition to all the participation in social networks and the simple searches for information, they are some of the sources that allow massive data collection.

These data linked to actions or daily activities are articulated with the most static data typical of personal information, such as personal identification number, date, and place of birth, affiliation, nationality, sex, status, civil, medical history, level of studies achieved, membership in sports or political associations, etc.

Using methodologies related to data mining, people's data sets are processed with the application of complex algorithms. As a result, accurate descriptions of personality characteristics and predictions about behavior and decisions are obtained.

The next step is the design of personalized messages, which are created based on the individual characteristics of each recipient so that " each citizen receives the most relevant information ."

Microtargeting Application in Marketing

Traditional marketing techniques focus on the product and target an undifferentiated market, with generic messages transmitted through traditional channels, such as radio or television. The paradigm shift focuses on the client/user and seeks to establish an empathic connection from the foreground of emotions, values, traditions, etc.

In that context, the techniques of microtargeting enable each customer or potential user to receive a specific advertising message, which was automatically selected from a list of different possible messages with selection criteria based on personal characteristics, wishes, and possibilities of the recipient.

Suggested articles:

Source