Success in books: a big data approach to bestsellers

Abstract

Reading remains the preferred leisure activity for most individuals, continuing to offer a unique path to knowledge and learning. As such, books remain an important cultural product, consumed widely. Yet, while over 3 million books are published each year, very few are read widely and less than 500 make it to the New York Times bestseller lists. And once there, only a handful of authors can command the lists for more than a few weeks. Here we bring a big data approach to book success by investigating the properties and sales trajectories of bestsellers. We find that there are seasonal patterns to book sales with more books being sold during holidays, and even among bestsellers, fiction books sell more copies than nonfiction books. General fiction and biographies make the list more often than any other genre books, and the higher a book's initial place in the rankings, the longer the book stays on the list as well. Looking at patterns characterizing authors, we find that fiction writers are more productive than nonfiction writers, commonly achieving bestseller status with multiple books. Additionally, there is no gender disparity among bestselling fiction authors but nonfiction, most bestsellers are written by male authors. Finally we find that there is a universal pattern to book sales. Using this universality we introduce a statistical model to explain the time evolution of sales. This model not only reproduces the entire sales trajectory of a book but also predicts the total number of copies it will sell in its lifetime, based on its early sales numbers. The analysis of the bestseller characteristics and the discovery of the universal nature of sales patterns with its driving forces are crucial for our understanding of the book industry, and more generally, of how we as a society interact with cultural products.

Success and luck in creative careers

“…Research in developmental psychology has studied careers of prominent artists and scientists for decades, advocating the importance of chance for the successful unfolding of careers in various creative domains [1,2,3,4]. In recent years, the availability of big databases on scientific publications [5] and artistic records, from books to movies [6,7,8], has made it possible to test a number of previously suggested hypotheses on a large scale. For instance, in previous work [9,10] the analysis of thousands of creative careers has shown that the biggest hit of an individual occurs randomly within an individual's career, a finding named the equal-odds-rule [3].…”

Success in books: predicting book sales before publication

“…In our previous work [16], we analyzed and modeled the dynamics of book sales, identifying a series of reproducible patterns: (i) most bestsellers reach their sales peak in less than ten weeks after release; (ii) sales follow a universal "early peak, slow decay" pattern that can be described by an accurate statistical model; (iii) we showed that the formula predicted by the model helps us predict future sales. Yet, to accurately predict the future sales using the model of Ref.…”

“…Yet, to accurately predict the future sales using the model of Ref. [16], we need at least the first 25 weeks of sales after publication, a period within which most books have already reached their peak sales and started to lose momentum. Therefore, predictions derived from this statistical model, potentially useful for long-term inventory management, are not particularly effective for foreseeing the sales potential of a new book.…”

The long-term impact of ranking algorithms in growing networks

“…Recent studies [37,56] emphasized the individuals' limited attention as a determinant for the viral popularity of low-quality items. Other recent works focused on modelling the interplay between quality/talent and popularity for diverse types of agents, including websites [20], scientific papers [30,55], researchers [29,50], and bestseller books [58]. Both model-based [20,30] and experimental results [43] indicate that in presence of social influence, the relation between popularity and quality is highly non-linear, meaning that a small variation of quality leads to large variations in popularity.…”