Organizational Ambidexterity Towards a Multilevel Understanding

Organizational Ambidexterity: Towards a

Multilevel Understanding

Zeki Simsek

University of Connecticut

abstract Although interest in the concept of organizational ambidexterity has increased

during recent years, this line of inquiry remains unfocused and limited due to a lack of more

encompassing conceptual efforts. As a first step, we begin by critically reviewing previous

research on the conceptualization, antecedents, and consequences of ambidexterity. We then

offer a model that specifies a more encompassing, multilevel explanation of ambidexterity.

Finally, we trace research and managerial implications and suggest some potential research

avenues.

INTRODUCTION

Increasingly, organizational researchers are using ambidexterity, the ability of humans to

use both hands with equal skill, as a metaphor for organizations that are equally

dexterous at exploiting and exploring. An ambidextrous organization maintains a high

degree of balance between exploitation (learning via local search, experiential refinement,

and reuse of existing knowledge) and exploration (learning gained through processes

of concerted variation, planned experimentation, and play) (Baum et al., 2000a;

Levinthal and March, 1993; March, 1991). Researchers have used ambidexterity to

analyse numerous significant organizational phenomena. Its importance has been noted

across the fields of strategic management ( Jansen et al., 2008; Lubatkin et al., 2006;

Smith and Tushman, 2005), innovation and technology management (Ambos et al.,

2008; He and Wong, 2004; Markman et al., 2008; Tushman and O'Reilly, 1996),

organizational learning and adaptation (Levinthal and March, 1993), organization

theory (Adler et al., 1999; Benner and Tushman, 2003), and organizational behaviour

(Gibson and Birkinshaw, 2004). Additionally, the managerial importance of ambidexterity

is evident in the many prescriptions offered for organizational performance

improvement, adaptation, and survival.

Address for reprints: Zeki Simsek, University of Connecticut, School of Business, Management Department,

2100 Hillside Road, Unit 1041, Storrs, CT 06269-1041, USA (Zeki.Simsek@Business.uconn.edu).

© Blackwell Publishing Ltd 2009. Published by Blackwell Publishing, 9600 Garsington Road, Oxford, OX4 2DQ, UK

and 350 Main Street, Malden, MA 02148, USA.

Journal of Management Studies 46:4 June 2009

doi: 10.1111/j.1467-6486.2009.00828.x

Curiously, however, organizational ambidexterity (OA) remains an undertheorized,

underconceptualized, and, therefore, poorly understood phenomenon. Perhaps Gupta

et al. (2006, p. 697) put it best when they noted that, 'although near consensus exists on

the need for balance [of exploitation and exploration], there is considerably less clarity

on how this balance can be achieved'. Importantly, research to date has typically

employed only one variable to explain OA, such as dual structures (Benner and

Tushman, 2003), behavioural context (Gibson and Birkinshaw, 2004), or top management

team (TMT) behavioural integration (Lubatkin et al., 2006). Lacking integrative

models spanning multiple levels of analysis (Gupta et al., 2006; Jansen et al., 2005),

previous studies thus far 'have not generated an overarching theory' (Adler et al., 1999)

to explain OA. Additionally, comparison, aggregation, and replication of findings have

been difficult because of diverse conceptualizations. Together, these issues highlight the

need for taking stock of the current body of knowledge, further specifying the OA

construct, and an encompassing model of OA.

As a first step, we review extant research and suggest that differentiating the various

input-process-output elements[1] associated with this line of inquiry can help invigorate

the OA concept by both substantiating and extending its conceptual foundations. Then,

we advance a multilevel explanatory model that concurrently extends and synthesizes

research by specifying the dominant relationships between constructs at the organizational,

interfirm, and environmental levels of analysis. By specifying these interactive

influences among these levels, we demonstrate that a thorough examination of such

influences may be critical in gaining a more complete understanding of where OA comes

from and how it matters to performance. In so doing, we respond to the call for more

integrative and multilevel analyses on OA (e.g. Gupta et al., 2006; Jansen et al., 2005;

Raisch and Birkinshaw, 2008). In particular, our theory addressees two neglected issues

that Raisch and Birkinshaw (2008) observe in a comprehensive review of this line of

research: (a) the interrelationships between different antecedents and (b) the complexity

of the ambidexterity-performance relationship. Finally, completing the conceptual