Which pathway to good ideas? An attention‐based view of innovation in social networks

• Author: Paul M. Leonardi,Luke Rhee
• Published date: 01 April 2018
• Date: 01 April 2018
• DOI: http://doi.org/10.1002/smj.2755

RESEARCH ARTICLE

Which pathway to good ideas? An attention-based

view of innovation in social networks

Luke Rhee

1

| Paul M. Leonardi

2

1

Department of Technology Management and

Innovation, Tandon School of Engineering,

New York University, Brooklyn, New York

2

Technology Management Program, College of

Engineering, University of California, Santa

Barbara, California

Correspondence

Luke Rhee, Department of Technology

Management and Innovation, Tandon School of

Engineering, New York University, Brooklyn, NY

11201.

Email: shyun.rhee@nyu.edu

Research Summary:This study introduces the notion of

attention allocation in networks to argue that individuals

with different types of network structure produce good

ideas via different pathways. Using survey dataon commu-

nication networks at a software company, we find that peo-

ple with highly constrained networks generate good ideas

by following a logic of interrogation, by which they focus

their attention on information from a particular contact.

Conversely, individuals with less constrained networks

produce good ideas by following a logic of recombination,

whereby they divide their attention to information coming

from across multiple contacts. The results show that in

highly constrained networks, interrogation is a more reli-

able pathway to good ideas than recombination. We dis-

cuss the implications of these findings for behavioral

strategy, social networks, and innovation.

Managerial Summary:People can develop good ideas

when they recombine diverse information inputs shared

by non-redundant communication partners that span mul-

tiple local clusters. But, in an organization, most individ-

uals are embedded in constrained networks of people

who know each other and thus typically receive redun-

dant information from work colleagues. This study sug-

gests that they can innovate via a different pathway:

through interrogation. We find that people who focus

their attention on information coming from a particular

person succeed at generating good ideas because they

deeply interrogate local knowledge and develop domain-

specific insights.

KEYWORDS

attention, behavioral strategy, good ideas, interrogation,

network structure, recombination

Received: 5 December 2016 Revised: 8 November 2017 Accepted: 14 November 2017 Published on: 23 January 2018

DOI: 10.1002/smj.2755

1188 Copyright © 2017 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/smj Strat Mgmt J. 2018;39:1188–1215.

1|INTRODUCTION

Scholars are increasingly arguing that innovations in product and process that help firms achieve

competitive advantage can be directly traced to ideas sourced from employees (Dahlander, O’Mah-

ony, & Gann, 2016; Furr, Christensen, & Dyer, 2014; Haas & Hansen, 2007). To understand which

employees are most likely to be able to generate good ideas,

1

scholars have turned to the study of

social networks. The findings of this line of work typically show the relationship between one’s

position in the firm’s communication network and his or her ability to generate good ideas

(Fleming, Mingo, & Chen, 2007; Tortoriello, 2015). Their theoretical arguments suggest that net-

work structures are both positively and negatively associated with one’s ability to come up with

good ideas because networks channel information (and it is from information that ideas are created)

and different network structures channel information in different ways (Aral & Van Alstyne, 2011).

Burt (1992) argues that the relationship between network structure and the production of good ideas

is best explained by focusing on network constraint. Constraint on an individual is high if his or her

communication partners normally talk directly to each other or if they transfer information indirectly

via a shared contact. People with less constrained networks talk to colleagues who do not directly

talk to each other and thus span more structural holes.

The premise underlying many studies linking network structure to good ideas is summarized by

Burt (2004, p. 356): “idea generation at some point involves someone moving knowledge from this

group to that, or combining bits of knowledge across groups.”In other words, good ideas are the

result of recombination with diverse information across from different domains (Fleming et al.,

2007; Hargadon & Sutton, 1997). The network structure best suited to the receipt of diverse infor-

mation, and ripe with possibility for engaging in recombination, is one in which an ego talks with

alters who themselves do not often share information with each other. As Burt (2004, p. 354)

described, people who span structural holes (hereafter, brokers) are positioned to see across different

domains more broadly than their peers because it is as if they use “over-the horizon radar in an air-

plane or an MRI in a medical procedure.”In this vein, recombination in less constrained networks,

made possible by access to non-redundant pockets of information, is typically seen as the primary

pathway to the generation of good ideas.

The result of this research is a view that while brokers are in the best position to generate good

ideas, employees with high network constraint have a difficult time doing so. When an individual

has a highly constrained network, the redundancy of information he or she receives tends to con-

strain his or her ability to create good ideas, even though high levels of constraint make it easy to

implement ideas, and garner support for them, after they are generated (Obstfeld, 2005; Reagans &

McEvily, 2003). However, if a firm’s goal is to enable as many employees as possible to contribute

good ideas, from the perspective of strategy the fact that only a few employees with a particular kind

of network structure are able to generate good ideas is problematic. It is difficult to become a broker

(Burt, 2002; Zaheer & Soda, 2009). Network research has shown that people have tendencies

toward homophily (Ibarra, 1992; McPherson, Smith-Lovin, & Cook, 2001) and triadic closure

(Granovetter, 1973), which results in overall cohesion (Coleman, 1990). Individuals often form ties

with others who are like them and who know their friends, such that they become embedded in

highly constrained networks of people who know each other. The result is that even people who find

1

Following the work of Amabile (1996) and Burt (2004), ideas are considered “good”when multiple expert observers within the

organization agree that they are. The “goodness”of ideas is in the eyes of the beholder because in any case, there are always some

external stakeholders that ought to evaluate an idea so that it is considered good. As will be explained in detail later in an empirical

sense, we utilize the consensus by expert observers to define ideas as “good.”

RHEE AND LEONARDI 1189