In the days before the Internet, retail marketing data was limited to those who used the business and the few that actively did not. Advertising was unidirectional, and nodal analysis stopped two spheres from the center. In those days, we were six degrees separated from anyone on Earth.
Facebook, Twitter, websites and review blogs have brought together the business and the customer. We are now four degrees separate. Where once we had hundreds of pieces of business data, now even a small business can have hundreds of thousands. The new business intelligence comes from the use of big data.
To understand big data, it is important to know the fundamentals of all data analysis. Data analysis is the process of applying statistical and logical evaluation to a group of information to condense and assess it. Typically, statistical analysis compares data points to the average, which gives it a bell curve or one of its variations. Or the data is predictive, giving it a regression line. Despite the size of the data set, all statistical analysis looks for the central tendency of the group.
Social Media’s Influence
Platforms like Facebook and Pinterest have given larger data sets. More importantly, the data is not static. In 2014, university researchers in collaboration with data managers at Facebook ran a big data experiment using nearly 700,000 people. The researchers tweaked the emotional content of the user’s feed and compared it to the type of information that the user placed on the feed. They found that positive news feed led to positive user contribution. The reverse was also true. Not only was this groundbreaking, due to the sheer volume of data, it was also a telling social psychology experiment. From a marketer’s perspective, the results are impressive. They show that positive business feeds create a tendency for positive customer responses. Likewise, no feed caused a withdrawal effect, leading to no social engagement.
Ultimately, data analysis finds its usefulness as a tool to predict shifts in central tendency based on the manipulation of one or more variables. The Facebook experiment controlled the emotion and looked at the result in the feed. The more data that can be collected, the more accurate the predictions. Offsite cloud storage systems give businesses of varying sizes the ability to collect business intelligence and analysis data that can be used to evaluate and predict consumer responses. Storage space can be scaled so that smaller businesses can hold the information they need without breaking their budget. Even simple analysis like averages and standard deviations, which show variation from the average, can be useful when a business has enough data.
Unlike advertising of the past, big data statistics are cyclical. The ability to learn is a hallmark of intelligence, which is one reason that business computer science people have coined the term business intelligence. One of the outcomes of big data analysis is a recursive ability to reintroduce adjusted information into the system and monitor the outcome. Using the implications of the Facebook study, a marketer will introduce emotional content into their advertising, analyze the results and change the advertising in response. This process is repeated and, if automated, gives the marketer a form of artificial intelligence focused on building a business.