Predictions are part of the human condition and have been made since the beginning of time. Getting it right is something else. New technologies bring greater reliability to surveys, but with a consequent margin of error. The analysis of big data, with the emergence of Big Data, helps predict success, but leaves open questions to which it will not be easy to find a solution.
Successes, at least in the world of creativity, are not achieved only by applying an algorithm that selects the necessary ingredients for the work to be applauded by everyone. However, it is proven that the combination of a series of elements, in the right proportions, will facilitate the success of a novel or any other product that wants to be put on the market.
Is it easy to make a “best seller”? Can it be guaranteed that a particular novel will be “best-selling” because its plot is developed following the guidelines set by an algorithm? Of course it can be achieved, but with a significant margin of error, which varies depending on the talent and creative capacity of its author. Otherwise, publishers would have banished the word “failure” from their vocabulary and would cease to believe in the intuition that made them gamble with greater or lesser fortune.
Two Stanford researchers-Jodie Archer and Matthew L. Jockers – recently presented an algorithm that guarantees, in 80% of cases, when a novel will be best-selling. The prediction of success is relative, although it is based on the comparative study of massive data – Big Data – properly analyzed and correctly selected to serve as a guideline for the author.
The history of literature is peppered with enormous errors, which have revealed the “smell” and “intuition” of some renowned editors. Carlos Barral, for example, never forgave himself for having rejected the manuscript of the novel “Cien años de soledad”, by Gabriel García Márquez, the day it was offered to him in Barcelona. Nor would editors who did not sense the success of J. K. Rowling’s “Harry Potter and the Philosopher’s Stone” or Stephen King’s “Carrie” forgive him.
It is possible that his tastes were not coincident with those of those promising authors, whom nobody knew then. Neither could they contrast those tastes with those of millions of potential readers, which is now possible, through the tools offered by Big Data. The information that today circulates through the market, through mobile phones, social networks and different Internet applications, is so extensive and abundant that the biggest problem lies in analyzing it and summarizing it with the greatest possible rigor and precision.
In the audiovisual sector, predictions are frequent, to the point of investing significant sums of money in analysing and developing them. Netflix, for example, invested a million dollars in the algorithm that identified the most attractive contents for potential viewers of the series “House of cards”. The success of this TELEVISION production has more than justified the investment. ” The contents of a series are no longer decided by a dozen people sitting at a table, but by the habits and behaviors of millions of viewers, “explained Joshua Lynn, president and co-founder of Piedmont Media Research, a company that uses algorithms to predict the success of productions before they are shot.
As David del Val, executive director of Telefónica I+D, pointed out in the second edition of the Impact Innovation Talks, “big data technologies are the key to developing artificial intelligence”.
Big Data specialists or, if you prefer, “experts in mass data processing”, are the profession of the future. What once seemed like a working hypothesis has become a tangible reality, although it does not achieve the desired effectiveness in all areas of life. For example, in politics, as demonstrated in the last American elections. The Democratic candidate, Hillary Clinton, opted for a campaign designed with the help of experts in Big Data, while the Republican Donald Trump was carried away more by the perception of his advisers and by the impressions collected in the electoral events.
In view of the election results, the eternal dilemma of whether quantity – an enormous volume of data – is better than quality – less but well-selected data-could arise when it comes to convincing undecided voters. When it comes to knowing which messages are most relevant to the electorate you’re targeting.
The famous three “uves” that identify Big Data-volume, speed and variety – could be summarized in the following motto:”more effective companies and more satisfied users”. Information grows 40% annually, but we cannot forget other factors, such as privacy.
Cost reduction, profitability, operational efficiency and personalized offering are in the DNA of Big Data, but there is still a long way to go.