A New Method for Identifying Recombinations of Existing Knowledge Associated with High-Impact Innovation* Satyam Mukherjee, Brian Uzzi, Ben Jones, and Michael Stringer How existing technologies and ideas are recombined into new innovations remains an important question, particularly as the store of prior technology, art, and work expands at an increasing rate. Yet, methodologies for identifying effective recombinations remain a nascent area of research. This paper extends our previous work, which developed a network methodology for assessing a scientific article’s recombinations of prior work. The methodology uses information from the entire co-citation network of all papers recorded in the Web of Science to identify combinations of prior work that are conventional or atypical and then identifies the virtuous mix of conventional and atypical pairings associated with high impact work. Here, we summarize our prior method and findings, present new findings, and perform a case study application to the field of management science. First, the results show that despite an ever-increasing frontier of possible new combinations of prior work, atypical combinations of prior work are becoming increasingly rare with time, while the distribution of conventional pairings is increasing with time. Second, our analyses show that with time the atypical pairings found in hit papers have a relatively stable mean rate at which they become conventional pairing. Nevertheless, the variance around the mean is growing significantly, which indicates that there is a greater tendency over time for novel pairings either to be virtually never used again or to become conventional pairings. Practitioner Points • High-impact work is sharply elevated when combina- tions of prior work are anchored in substantial conven- tionality while mixing in a left tail of combinations that are rarely seen together. • In business, companies and analysts could better predict the value of a future patent or product based on its combination of high conventionality and infusions of novelty. • Consumer products like computer games that are able to incorporate conventional pairings with a small compo- nent of novelty may be more likely to be popular. Introduction I nvention is spurred on when existing innovations or their components are assembled into original new designs (Becker, 1982; Guimera, Uzzi, Spiro, and Amaral, 2005; Jones, 2009; Jones, Wuchty, and Uzzi, 2008; Schilling, 2005; Schumpeter, 1939; Usher, 1954; Uzzi and Spiro, 2005; Weitzman, 1998). Nevertheless, it remains a mystery as to what recombined elements of existing innovations or components are likely to be prom- ising and attractive to their intended audiences. In some creative fields, it is often lamented that it is impossible to know which recombinations will be fruitful (Committee on Facilitating Interdisciplinary Research, 2004). In film- making, it is accepted wisdom that “all hits are flukes,” an idiom echoing Samuel Goldwin’s quip that “nobody knows nothing” about what combination of previously used, or altogether new plot details, actors, settings, or direct references to past movies makes for a hit (Spitz and Horvát, 2014a, 2014b). As a consequence, many expen- sive failures must be suffered in the hope that one hit can recover the costs of experimentation. To some degree, similar uncertainty lies behind innovation in general (Börner et al., 2011; Evans and Foster, 2011; Fiore, 2008; Stokols, Hall, Taylor, and Moser, 2008; Wuchty, Jones, and Uzzi, 2007). Nevertheless, innovative ideas can be difficult to absorb (Henderson and Clark, 1990) and communicate as well, leading creatives to intentionally display conven- tionality (familiar combinations used many times in pre- vious work) side-by-side with atypical combinations. In his Principia, Newton presented his laws of gravitation using accepted geometry rather than his newly developed calculus despite the latter’s importance in developing his Address correspondence to: Satyam Mukherjee, Kellogg School of Management and Northwestern Institute on Complex Systems (NICO), 600 Foster Street, Evanston, IL 60208. E-mail: satyam.mukherjee@gmail.com. Tel: 847-491-2527. * Sponsored by the Northwestern University Institute on Complex Systems (NICO), by the Army Research Laboratory under Cooperative Agreement Number W911NF-09-2-0053, and DARPA BAA-11-64, Social Media in Strategic Communication. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the Army Research Laboratory or the U.S. Government. J PROD INNOV MANAG 2015;••(••):••–•• © 2015 Product Development & Management Association DOI: 10.1111/jpim.12294