Abhishek Nagaraj
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Abhishek Nagaraj has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Mathijs de Vaan
University of California Berkeley
2 shared publications Arul Murugan
School of Information University of California, Berkeley Berkeley California USA
1 shared publicationCecil‐Francis Brenninkmeijer
Fuqua School of Business Duke University Durham North Carolina USA
1 shared publication- Esther Shears
University of California, Berkeley, California, United States of America
1 shared publication - Gauri Subramani
Management Department, Lehigh University, Bethlehem, Pennsylvania 18015
1 shared publication - Henning Piezunka
The Wharton School, University of Pennsylvania, Philadelphia, Pennsylvania 19104
1 shared publication
Latest Publications
- Open Access
- 29 July 2026
- Published by Wiley in Strategic Management Journal
- Vol. 47 (11) , 2932-2958
Simulating strategic interactions with AI agents
Research Summary: We explore how Large Language Models (LLMs) can serve as synthetic subjects to inform strategy research. We introduce a framework for designing and running simulated experiments with LLM‐powered agents. We argue that this approach is useful for rapid, low‐cost prototyping of human experiments and for generating novel hypotheses. We apply the framework to the exploration–exploitation dilemma and show that LLM‐based experiments reproduce patterns observed among human participants. We then vary parameters and boundary conditions to illustrate how the same setup can support design iteration and surface hypotheses about when and why established results change. In the conclusion, we discuss the promise and limitations of artificial intelligence agents as “model organisms” for strategy. Managerial Summary: Artificial intelligence (AI) agents are beginning to enter firms as tools that can execute work, from writing code to coordinating complex tasks across systems. This article argues that their value for strategy extends beyond task automation: AI agents can also be used to simulate strategic interactions and assess how strategies perform under alternative assumptions. In our exploration–exploitation application, these simulations reproduce core patterns from prior human experiments and reveal where those patterns weaken or reverse. Used this way, AI agents can help firms prototype strategic choices, stress‐test assumptions, and direct managerial attention toward promising leads before larger commitments of time and effort.
- 29 July 2026
- Published by Wiley in Strategic Management Journal
- Vol. 47 (11) , 2932-2958
Open AccessSimulating strategic interactions with AI agents
Research Summary: We explore how Large Language Models (LLMs) can serve as synthetic subjects to inform strategy research. We introduce a framework for designing and running simulated experiments with LLM‐powered agents. We argue that this approach is useful for rapid, low‐cost prototyping of human experiments and for generating novel hypotheses. We apply the framework to the exploration–exploitation dilemma and show that LLM‐based experiments reproduce patterns observed among human participants. We then vary parameters and boundary conditions to illustrate how the same setup can support design iteration and surface hypotheses about when and why established results change. In the conclusion, we discuss the promise and limitations of artificial intelligence agents as “model organisms” for strategy. Managerial Summary: Artificial intelligence (AI) agents are beginning to enter firms as tools that can execute work, from writing code to coordinating complex tasks across systems. This article argues that their value for strategy extends beyond task automation: AI agents can also be used to simulate strategic interactions and assess how strategies perform under alternative assumptions. In our exploration–exploitation application, these simulations reproduce core patterns from prior human experiments and reveal where those patterns weaken or reverse. Used this way, AI agents can help firms prototype strategic choices, stress‐test assumptions, and direct managerial attention toward promising leads before larger commitments of time and effort.
- 23 February 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 72 (10) , 9199-9214
Getting on the Map: The Impact of Online Listings on Business Performance
We evaluate the extent to which small businesses maintain an online presence and estimate the impact of digital representation on business performance. Looking at a review platform, we find that roughly 18% of bars and restaurants in our sample do not have a listing as of the end of 2017, even though creating one is free. We then estimate the effect of adding a new listing using tax records on revenues combined with temporal variation in listing activity as well as a natural experiment that added over a thousand businesses to the platform via a data acquisition. The results suggest that new listings increase revenues in the range of 5%–10%; however, these effects vary substantially across establishments. Establishments that struggle to develop a reputation (e.g., nonchains or those in tourist areas) and those with a higher quality of product (e.g., larger restaurants or those with good reviews) experience larger increases after being listed. This paper was accepted by Duncan Simester, marketing. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2024.04716 .
- 23 February 2026
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 72 (10) , 9199-9214
Getting on the Map: The Impact of Online Listings on Business Performance
We evaluate the extent to which small businesses maintain an online presence and estimate the impact of digital representation on business performance. Looking at a review platform, we find that roughly 18% of bars and restaurants in our sample do not have a listing as of the end of 2017, even though creating one is free. We then estimate the effect of adding a new listing using tax records on revenues combined with temporal variation in listing activity as well as a natural experiment that added over a thousand businesses to the platform via a data acquisition. The results suggest that new listings increase revenues in the range of 5%–10%; however, these effects vary substantially across establishments. Establishments that struggle to develop a reputation (e.g., nonchains or those in tourist areas) and those with a higher quality of product (e.g., larger restaurants or those with good reviews) experience larger increases after being listed. This paper was accepted by Duncan Simester, marketing. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2024.04716 .
- 1 September 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Strategy Science
- Vol. 9 (3) , 277-296
The Divergent Effect of Competition on Platforms: Deterring Recruits, Motivating Converts
Knowledge platforms differ in how they source their knowledge; they can be categorized as traditional (e.g., Encyclopedia Britannica, Guide Michelin) or crowdsourced (e.g., Wikipedia, Yelp). Although research has compared and contrasted the two, we study how they compete with one another. In particular, we focus on how competition from traditional platforms affects volunteers’ contributions to crowdsourced platforms. We propose a divergent effect: competition increases individual contributions from volunteers who were already active before the entry of any competitor (pre-entry contributors). However, competition also hinders recruitment and therefore leads to a decrease in total contributions from the set of volunteers who are new to the platform in a given time period (newly recruited contributors). We suggest the first effect is driven by pre-entry contributors’ ideological attachment, whereas the second is driven by lower awareness of the platform postcompetitive entry. To test our theory, we leverage and examine how the phased entry of Google Maps in different countries affects contributions to OpenStreetMap, a crowdsourced mapping platform. We find the hypothesized divergent effects and provide suggestive evidence of the underlying mechanisms. We contribute to research on crowdsourcing, volunteers’ willingness to contribute, platform competition, and on the effect of competition on organizations more broadly. Funding: The authors acknowledge the financial support of the INSEAD Alumni Fund. Supplemental Material: The online appendices are available at https://doi.org/10.1287/stsc.2022.0125 .
- 1 September 2024
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Strategy Science
- Vol. 9 (3) , 277-296
The Divergent Effect of Competition on Platforms: Deterring Recruits, Motivating Converts
Knowledge platforms differ in how they source their knowledge; they can be categorized as traditional (e.g., Encyclopedia Britannica, Guide Michelin) or crowdsourced (e.g., Wikipedia, Yelp). Although research has compared and contrasted the two, we study how they compete with one another. In particular, we focus on how competition from traditional platforms affects volunteers’ contributions to crowdsourced platforms. We propose a divergent effect: competition increases individual contributions from volunteers who were already active before the entry of any competitor (pre-entry contributors). However, competition also hinders recruitment and therefore leads to a decrease in total contributions from the set of volunteers who are new to the platform in a given time period (newly recruited contributors). We suggest the first effect is driven by pre-entry contributors’ ideological attachment, whereas the second is driven by lower awareness of the platform postcompetitive entry. To test our theory, we leverage and examine how the phased entry of Google Maps in different countries affects contributions to OpenStreetMap, a crowdsourced mapping platform. We find the hypothesized divergent effects and provide suggestive evidence of the underlying mechanisms. We contribute to research on crowdsourcing, volunteers’ willingness to contribute, platform competition, and on the effect of competition on organizations more broadly. Funding: The authors acknowledge the financial support of the INSEAD Alumni Fund. Supplemental Material: The online appendices are available at https://doi.org/10.1287/stsc.2022.0125 .
- Open Access
The Economics of Augmented and Virtual Reality
This paper explores the economics of Augmented Reality (AR) and Virtual Reality (VR) technologies within decision-making contexts. Two metrics are proposed: Context Entropy, the informational complexity of an environment, and Context Immersivity, the value from full immersion. The analysis suggests that AR technologies assist in understanding complex contexts, while VR technologies provide access to distant, risky, or expensive environments. The paper provides a framework for assessing the value of AR and VR applications in various business sectors by evaluating the pre-existing context entropy and context immersivity. The goal is to identify areas where immersive technologies can significantly impact and distinguish those that may be overhyped.
Open AccessThe Economics of Augmented and Virtual Reality
This paper explores the economics of Augmented Reality (AR) and Virtual Reality (VR) technologies within decision-making contexts. Two metrics are proposed: Context Entropy, the informational complexity of an environment, and Context Immersivity, the value from full immersion. The analysis suggests that AR technologies assist in understanding complex contexts, while VR technologies provide access to distant, risky, or expensive environments. The paper provides a framework for assessing the value of AR and VR applications in various business sectors by evaluating the pre-existing context entropy and context immersivity. The goal is to identify areas where immersive technologies can significantly impact and distinguish those that may be overhyped.
- Top Cited Papers
- 1 January 2022
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 68 (1) , 564-582
The Private Impact of Public Data: Landsat Satellite Maps Increased Gold Discoveries and Encouraged Entry
How does public data shape the relative performance of incumbents and entrants in the private sector? Using a simple theoretical framework, I argue that public data reduces investment uncertainty, facilitates the discovery of new market opportunities, and increases the relative market share of new entrants relative to incumbents. I shed light on these predictions by estimating the impact of public data from Landsat, a U.S. National Aeronautics and Space Administration satellite mapping program, on the discovery rates of new deposits by incumbents (seniors) and entrants (juniors) in the gold exploration industry. I exploit idiosyncratic timing variation and cloud cover in Landsat coverage across regions to identify the causal effect of public data on the patterns of gold discovery. I find that Landsat data nearly doubled the rate of significant gold discoveries after a region was mapped and increased the market share of new entrants from about 10% to 25%. Public data seem to play an important, yet relatively underexplored, role in driving performance differences across firms. This paper was accepted by Ashish Arora, entrepreneurship and innovation.
- 1 January 2022
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 68 (1) , 564-582
Top Cited PapersThe Private Impact of Public Data: Landsat Satellite Maps Increased Gold Discoveries and Encouraged Entry
How does public data shape the relative performance of incumbents and entrants in the private sector? Using a simple theoretical framework, I argue that public data reduces investment uncertainty, facilitates the discovery of new market opportunities, and increases the relative market share of new entrants relative to incumbents. I shed light on these predictions by estimating the impact of public data from Landsat, a U.S. National Aeronautics and Space Administration satellite mapping program, on the discovery rates of new deposits by incumbents (seniors) and entrants (juniors) in the gold exploration industry. I exploit idiosyncratic timing variation and cloud cover in Landsat coverage across regions to identify the causal effect of public data on the patterns of gold discovery. I find that Landsat data nearly doubled the rate of significant gold discoveries after a region was mapped and increased the market share of new entrants from about 10% to 25%. Public data seem to play an important, yet relatively underexplored, role in driving performance differences across firms. This paper was accepted by Ashish Arora, entrepreneurship and innovation.
- 1 August 2021
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 67 (8) , 4908-4934
Information Seeding and Knowledge Production in Online Communities: Evidence from OpenStreetMap
The wild success of a few online communities (such as Wikipedia) has obscured the fact that most attempts at forming such communities fail. This study evaluates information seeding, an early-stage intervention to bootstrap online communities that enables contributors to build on externally sourced information rather than have them start from scratch. I analyze the effects of information seeding on follow-on contributions using data on more than 350 million contributions made by more than 577,000 contributors to OpenStreetMap, a crowd-sourced map-making community seeded with data from the U.S. Census. I estimate the effect of seeding using a natural experiment in which an oversight caused about 60% of U.S. counties to be seeded with a complete census map, while the rest were seeded with less complete versions. Although access to basic knowledge generally encourages downstream knowledge production, I find that a higher level of information seeding significantly lowered follow-on contributions and contributor activity on OpenStreetMap, and was associated with lower levels of long-term quality. However, seeding did benefit densely populated urban areas and did not discourage more committed users. To explain these patterns, I argue that information seeding can crowd out contributors’ ability to develop ownership over baseline knowledge and thereby disincentivize follow-on contributions. This paper was accepted by Chris Forman, information systems.
- 1 August 2021
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 67 (8) , 4908-4934
Information Seeding and Knowledge Production in Online Communities: Evidence from OpenStreetMap
The wild success of a few online communities (such as Wikipedia) has obscured the fact that most attempts at forming such communities fail. This study evaluates information seeding, an early-stage intervention to bootstrap online communities that enables contributors to build on externally sourced information rather than have them start from scratch. I analyze the effects of information seeding on follow-on contributions using data on more than 350 million contributions made by more than 577,000 contributors to OpenStreetMap, a crowd-sourced map-making community seeded with data from the U.S. Census. I estimate the effect of seeding using a natural experiment in which an oversight caused about 60% of U.S. counties to be seeded with a complete census map, while the rest were seeded with less complete versions. Although access to basic knowledge generally encourages downstream knowledge production, I find that a higher level of information seeding significantly lowered follow-on contributions and contributor activity on OpenStreetMap, and was associated with lower levels of long-term quality. However, seeding did benefit densely populated urban areas and did not discourage more committed users. To explain these patterns, I argue that information seeding can crowd out contributors’ ability to develop ownership over baseline knowledge and thereby disincentivize follow-on contributions. This paper was accepted by Chris Forman, information systems.
- 1 July 2021
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 67 (7) , 4446-4454
Social Learning in the COVID-19 Pandemic: Community Establishments’ Closure Decisions Follow Those of Nearby Chain Establishments
As conveners that bring various stakeholders into the same physical space, firms can powerfully influence the course of pandemics such as coronavirus disease 2019. Even when operating under government orders and health guidelines, firms have considerable discretion to keep their establishments open or closed during a pandemic. We examine the role of social learning in the exercise of this discretion at the establishment level. In particular, we evaluate how the closure decisions of chain establishments, which are associated with national brands, affect those of proximate, same-industry community establishments, which are independently owned or managed. We conduct these analyses using cell phone location tracking data on daily visits to 230,403 U.S.-based community establishments that are colocated with chain establishments affiliated with 319 national brands. We disentangle the effects of social learning from confounding factors by using an instrumental variables strategy that relies on local variation in community establishments’ exposure to closure decisions made by brands at the national level. Our results suggest that closing decisions of community establishments are affected by the decisions made by chain establishments; a community establishment is 3.5% more likely to be open on a given day if the proportion of nearby open chain establishments increases by one standard deviation. This paper was accepted by Olav Sorenson, organizations.
- 1 July 2021
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 67 (7) , 4446-4454
Social Learning in the COVID-19 Pandemic: Community Establishments’ Closure Decisions Follow Those of Nearby Chain Establishments
As conveners that bring various stakeholders into the same physical space, firms can powerfully influence the course of pandemics such as coronavirus disease 2019. Even when operating under government orders and health guidelines, firms have considerable discretion to keep their establishments open or closed during a pandemic. We examine the role of social learning in the exercise of this discretion at the establishment level. In particular, we evaluate how the closure decisions of chain establishments, which are associated with national brands, affect those of proximate, same-industry community establishments, which are independently owned or managed. We conduct these analyses using cell phone location tracking data on daily visits to 230,403 U.S.-based community establishments that are colocated with chain establishments affiliated with 319 national brands. We disentangle the effects of social learning from confounding factors by using an instrumental variables strategy that relies on local variation in community establishments’ exposure to closure decisions made by brands at the national level. Our results suggest that closing decisions of community establishments are affected by the decisions made by chain establishments; a community establishment is 3.5% more likely to be open on a given day if the proportion of nearby open chain establishments increases by one standard deviation. This paper was accepted by Olav Sorenson, organizations.
- Top Cited Papers
- 8 September 2020
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 117 (38) , 23490-23498
Improving data access democratizes and diversifies science
Significance: Data access is critical to empirical research, but past work on open access is largely restricted to the life sciences and has not directly analyzed the impact of data access restrictions. We analyze the impact of improved data access on the quantity, quality, and diversity of scientific research. We focus on the effects of a shift in the accessibility of satellite imagery data from Landsat, a NASA program that provides valuable remote-sensing data. Our results suggest that improved access to scientific data can lead to a large increase in the quantity and quality of scientific research. Further, better data access disproportionately enables the entry of scientists with fewer resources, and it promotes diversity of scientific research.
- 8 September 2020
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 117 (38) , 23490-23498
Top Cited PapersImproving data access democratizes and diversifies science
Significance: Data access is critical to empirical research, but past work on open access is largely restricted to the life sciences and has not directly analyzed the impact of data access restrictions. We analyze the impact of improved data access on the quantity, quality, and diversity of scientific research. We focus on the effects of a shift in the accessibility of satellite imagery data from Landsat, a NASA program that provides valuable remote-sensing data. Our results suggest that improved access to scientific data can lead to a large increase in the quantity and quality of scientific research. Further, better data access disproportionately enables the entry of scientists with fewer resources, and it promotes diversity of scientific research.
- 1 July 2018
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 64 (7) , 3091-3107
Does Copyright Affect Reuse? Evidence from Google Books and Wikipedia
While digitization has greatly increased the reuse of knowledge, this study shows how these benefits might be mitigated by copyright restrictions. I use the digitization of in-copyright and out-of-copyright issues of Baseball Digest magazine by Google Books to measure the impact of copyright on knowledge reuse in Wikipedia. I exploit a feature of the 1909 Copyright Act whereby material published before 1964 has lapsed into the public domain, allowing for the causal estimation of the impact of copyright across this sharp cutoff. I find that, while digitization encourages knowledge reuse, copyright restrictions reduce citations to copyrighted issues of Baseball Digest by up to 135% and affect readership by reducing traffic to affected pages by 20%. These impacts are highly uneven: copyright hurts the reuse of images rather than text and affects Wikipedia pages for less-popular players greater than more-popular ones.The online appendix is available at https://doi.org/10.1287/mnsc.2017.2767 .This paper was accepted by Lee Fleming, entrepreneurship and innovation.
- 1 July 2018
- Published by Institute for Operations Research and the Management Sciences (INFORMS) in Management Science
- Vol. 64 (7) , 3091-3107
Does Copyright Affect Reuse? Evidence from Google Books and Wikipedia
While digitization has greatly increased the reuse of knowledge, this study shows how these benefits might be mitigated by copyright restrictions. I use the digitization of in-copyright and out-of-copyright issues of Baseball Digest magazine by Google Books to measure the impact of copyright on knowledge reuse in Wikipedia. I exploit a feature of the 1909 Copyright Act whereby material published before 1964 has lapsed into the public domain, allowing for the causal estimation of the impact of copyright across this sharp cutoff. I find that, while digitization encourages knowledge reuse, copyright restrictions reduce citations to copyrighted issues of Baseball Digest by up to 135% and affect readership by reducing traffic to affected pages by 20%. These impacts are highly uneven: copyright hurts the reuse of images rather than text and affects Wikipedia pages for less-popular players greater than more-popular ones.The online appendix is available at https://doi.org/10.1287/mnsc.2017.2767 .This paper was accepted by Lee Fleming, entrepreneurship and innovation.
- 1 August 2017
- Published by Academy of Management in Academy of Management Proceedings
- Vol. 2017 (1) , 13110
Does Open Data Spur Online Communities? Evidence from Crowdsourced Mapping
Does public provision of basic knowledge spur follow-on community-based innovation? I analyze the impact of the US Census TIGER mapping project, a 8-year long public effort that made available a complete digital map of the United States on OpenStreetMap, a Wikipedia-style community-contributed mapping project used widely on the internet. As a result of an error, when TIGER was first used in OpenStreetMap about 60% of quasi-randomly chosen counties received public investments in new knowledge, while the others did not. By comparing user activity in regions that both received and did not receive public information in a difference-in-difference framework, I estimate the impact of the TIGER project on the OpenStreetMap community. I find that public investments in new knowledge hurts the development of the OpenStreetMap community–both by reducing the number of contributions and users and also by depressing the production of follow-on knowledge. Publicly-provided knowledge hurts the community, not by reducing the arrival of new users, but rather by failing to convert new users into committed contributors.
- 1 August 2017
- Published by Academy of Management in Academy of Management Proceedings
- Vol. 2017 (1) , 13110
Does Open Data Spur Online Communities? Evidence from Crowdsourced Mapping
Does public provision of basic knowledge spur follow-on community-based innovation? I analyze the impact of the US Census TIGER mapping project, a 8-year long public effort that made available a complete digital map of the United States on OpenStreetMap, a Wikipedia-style community-contributed mapping project used widely on the internet. As a result of an error, when TIGER was first used in OpenStreetMap about 60% of quasi-randomly chosen counties received public investments in new knowledge, while the others did not. By comparing user activity in regions that both received and did not receive public information in a difference-in-difference framework, I estimate the impact of the TIGER project on the OpenStreetMap community. I find that public investments in new knowledge hurts the development of the OpenStreetMap community–both by reducing the number of contributions and users and also by depressing the production of follow-on knowledge. Publicly-provided knowledge hurts the community, not by reducing the arrival of new users, but rather by failing to convert new users into committed contributors.