Selected Work-in-progress (expand for abstracts)
Intra-firm collaboration, technological network centrality, and their varied influence on employees’ probability of intra-firm promotions.
Bruno Cirillo, Daniel Tzabbar, Stefano Breschi.
The study delves into the complex landscape of promotions within a specific group: research and development (R&D) engineers functioning as inventors in corporate settings. By investigating the influence of key central positions—specifically, collaborative and technological centrality—within intra-company networks on engineers' likelihood of promotion across different ranks, it aims to challenge established assumptions. Analyzing the career paths and patenting histories of 4,416 engineers at a leading U.S. semiconductor company from 1993 to 2012, we unveil nuanced insights. While the prevailing assumption suggests that central positions enhance promotion opportunities across mid-levels, the findings bring forth an innovative insight. Collaborative centrality notably surpasses technological centrality in boosting promotional prospects across all ranks, contrary to initial beliefs. Interestingly, technological centrality, beneficial in lower to mid-level positions, takes on a different role at higher ranks, acting as a hindrance that diminishes the chances of advancement. This study offers significant insights into the varying effects of different centralities on promotion opportunities among R&D engineers, challenging entrenched beliefs about the differential impacts of distinct centrality types within organizational hierarchies.
Who benefits from knowledge spill-ins from corporate investments in spinouts?
Bruno Cirillo, Valentina Fani, Dennis Park.
Established companies often invest in startups through corporate venture capital (CVC) to harness knowledge spill-ins, bolstering their own innovation. Yet, they anticipate a potential sharing of their knowledge with other established firms via their startup investments. Notably, close social ties between startup founders and their parent companies can facilitate the transfer of CVC investors’ knowledge to the parent firms. Our study investigates the changing dynamics of knowledge flows between CVC investors and the parent companies of their portfolio firms over time. We found that immediately after CVC investments, parent firms tend to experience increased knowledge inflows from CVC investors due to the advantages stemming from human and social capital shared with startup founders. However, over time, this trend reverses, with CVC investors benefiting from greater knowledge inflows from parent companies as their relationships with startups solidify and their grasp of the parent firms’ knowledge deepens. These findings highlight an intricate evolutionary facet of knowledge exchange between CVC investors and parent firms, emphasizing the role of startup human and social capital as intermediaries in these knowledge flows.
Can success arrive too soon? The effect of early initial success on venture’s innovation behavior in terms of exploitation and exploration.
Arusyak Zakarian, Stefano Breschi, Bruno Cirillo, Daniel Tzabbar.
The influence of a venture's early success, along with its potential negative repercussions on future growth and success, is an underexamined area in understanding organizational behavior and future innovation. This research aims to explore how initial success influences organizational learning processes and subsequent innovation behaviors in young medical device firms within the United States, specializing in Cardiovascular, Orthopedic, and Radiology fields. Leveraging an original dataset sourced from FDA records, Crunchbase, LinkedIn profiles, PatentsView, and web searches, the study investigates the effects of early success. It uncovers that while early success positively influences subsequent exploitative product success, it detrimentally affects exploratory product success. Additionally, the research examines the moderating roles of founding team venturing experience and the breadth of inventors' knowledge. This in-depth analysis provides essential insights into the intricate challenges faced by young ventures, emphasizing the delicate balance between early success for survival and its potential long-term impact on innovation performance.
The interplay between individual experience and Artificial Intelligence in predicting firm’s performance
Artyom Yepremyan, Francesco Castellaneta, Bruno Cirillo.
The rapid advancement of Artificial Intelligence (AI) has significantly transformed decision-making processes across various sectors. While firms increasingly depend on AI's predictive abilities, its effectiveness can face obstacles due to factors such as data scarcity and biases. Understanding the interplay between human experience and AI in decision-making is crucial. Some studies underscore the value of accumulated experience in improving forecasting, while others highlight potential biases associated with experience accumulation, potentially hampering adaptability and accuracy.
This study delves into how specific and general individual experiences impact AI's predictive abilities in forecasting firm performance. To do this, we built an extensive dataset comprising financial analysts and NYSE-listed firms from 1983 to 1998. We employed Random Forest (RF) models and individual forecasts to compare their predictive performances and understand how individual experience can train AI models for improved performance. Findings reveal that individuals with extensive firm-specific experience tend to exhibit higher errors in predicting earnings-per-share (EPS) compared to AI algorithms, particularly in uncertain environments. Conversely, individuals with higher levels of general experience displayed reduced errors. Additionally, increased uncertainty amplified the importance of both specific and general experiences, mitigating biases associated with experience accumulation. Furthermore, high firm-specific experience in the training set did not result in higher RF forecasting errors, while higher general experience reduced errors. Both specific and general experiences decreased RF error in higher uncertainty situations. However, in scenarios with increased information asymmetry, higher firm-specific experience in the training set led to increased RF error.
These results underscore the positive impact of diverse experiences, especially under uncertainty, on decision-making and the integration of AI. Specific experiences add nuances, while general experiences broaden perspectives for AI. Nevertheless, high information asymmetry may limit AI's accuracy when predominantly relying on specific experiences. Understanding these dynamics forms a basis for leveraging human experiences to enhance AI capabilities in decision-making contexts.