Chair: Vish Krishnan

Paper 1: Marketing, R&D and Valuation of Startup Firms, by Joglekar & Lévesque

The problem focus is on startups with staged venture financing, where both R&D and marketing are significant fractions of overall expenses. When should a manager of such a startup firm acquire working capital and how should she distribute that capital between R&D ­ thus improving product quality ­ and marketing expenses ­ thus building sales up in order to grow valuation? Also, should there be a cap on the total R&D and marketing spending to increase profitability during such staged venture financing? We develop a model to study resource acquisition and allocation across successive stages of startup growth. Model analysis provides insights on optimal acquisition and allocation policies and characterizes the impact of changes in productivity and rate of evolution of the R&D and marketing payoff on the underlying decision problems. Our results also illustrate conditions on optimal capping of R&D and marketing expenses as a percentage of revenues.

Paper 2: Innovation, Openness, and Platform Control Parker & Van Alstyne [PDF]

In markets characterized by platforms and their ecosystems of complementary applications, we find that competition among application developers can reduce innovation while competition among platforms can increase innovation. Developers can be better off submitting to platform control as opposed to producing for an unsponsored platform. Although a social planner would open a platform sooner and to a greater degree than would a private platform sponsor, a platform sponsor's ability to control downstream innovation gives it reason to behave more like a social planner. However, if platforms are to perform this role, platform sponsors need longer duration rights than application developers. Results can inform antitrust and intellectual property regulation, technological innovation, competition policy, and intellectual property strategy.

Chair: Geoffrey Parker

Paper 1: An Evaluation of Analytical, Simulation, and Statistical Concurrent Engineering Models Gerwin, Bhuiyan, & Thomson

We suggest directions for new product development (NPD) research by quantitatively evaluating theories associated with different methodologies. The focus is on concurrent engineering models that use analytical, simulation or statistical methods. Through a new approach that collects data from 40 research articles, we determine the proportion of models based on a given methodology that have incorporated each of certain significant NPD features such as overlapping. A trend analysis based on logistic regression indicates that, with at most one exception, the various proportions have not changed over time. Next, in an analysis of the features appearing simultaneously in a majority of a methodologys models, we find that there are stable patterns of features. We then recommend features that researchers using each method should consider. Our study also uncovers beneficial knowledge that researchers associated with different methods can learn from each other.

Paper 2: The Impact of Component Modularity on Design Evolution: Evidence from the Software Industry, MacCormack, Rusnak & Baldwin

Much academic work asserts a relationship between the design of a complex system and the manner in which this system evolves over time. In particular, designs that are more modular are argued to be more “evolvable,” in that they facilitate making future adaptations, the nature of which do not have to be specified in advance. In essence, modularity creates option value with respect to new and improved designs. We explore this phenomenon in software by analyzing the evolution of a successful commercial product. We find measures of modularity predict three different aspects of component evolution: survival, maintainability and augmentation. Specifically, we show that tightly-coupled components are more likely to survive between versions, more likely to require maintenance, and less likely to be added over time. The results have major implications for the designers of systems that must meet uncertain future demands.

Chair: Janice Carillo

Paper 1: The Incumbent’s Fate in the Face of Innovation, Schmidt & Porteus

The literature suggests an incumbent is susceptible to being displaced by an entrant in the face of an innovation that is competence-destroying, drastic in magnitude (i.e., radical), and/or disruptive. Another possible factor is a lack of complementary assets in marketing and manufacturing. Our model simultaneously examines the impact of all four factors, finding: 1) In overcoming the replacement effect, complementary assets are all that matter if the innovation is "barely incremental," 2) The incumbent will always invest in a "highly drastic" competence-enhancing innovation, and also in a competence-destroying one if it has a positive level of complementary assets; domination requires both; 3) Domination always requires some minimal level of complementary assets; 4) The incumbent may actually prefer a radical innovation to an incremental one; 5) to deter the entrant's introduction of an inferior product, the incumbent needs to be able to produce its own version so cheaply that it sees its new product as an opportunity rather than a threat:; and 6) The complementary asset level required for an incumbent to capitalize on disruptive innovation is no higher than for the sustaining type, given an equivalent competence enhancing/ destroying parameter.

Paper 2: The Hidden Perils of Career Concerns in R&D Organizations: Experimental Evidence; Siemsen & Katok

We use laboratory experiments to test how players with career concerns choose the difficulty of their organizational tasks. Drawing upon existing theory, we subject participants in our experiments to a context in which they act as agents for a principal, and have to convince this principal of their capability in order reap financial incentives. To do so, they can increase the difficulty of their assigned task, thereby lowering their likelihood of obtaining a successful outcome. Theory suggests that the difficulty of the task influences their expected gain in reputation. The data from the experiment supports that people have a higher propensity to choose a difficult solution as the expected reputation they can gain by choosing this solution increases. We compare the participants' behavior in conditions with and without such career concerns. Further, we show that highly capable and less capable agents have different strategies as long as their choice is unobservable. However, the less capable agents tend to pool on the highly capable agents' strategy as their choice becomes observable by the principal. We also show that performance rewards reduce the less-capable agents' propensity to choose the difficult task. Interestingly, though, the highly capable agents choice pattern is unaffected by performance rewards.

Chair: Nitin Joglekar

Paper 1: Managing Development Problem Solving Using Delegated Search: The Role of Design Specification and Awards, Erat and Krishnan

We study the emerging trend of firms delegating the search for solutions to their unstructured design problems to groups of outside agents. One of the key benefits from such an approach is the variety of solution approaches that may be pursued (in parallel) by these agents. Delegated search, while offering benefits such as lower costs and/or ability to search over large parts of the solution space, also poses significant challenges depending on the problem type.   We examine two related research questions in this paper: (i) How should a firm manage their problem specification before broadcasting it to the outside experts, and how does the extent of problem specification affect the breadth of solution being searched?, and (ii) How does the award structure (i.e., number and size of prizes/awards) affect the breadth of the solution space that will be searched?

We identify a robust phenomenon—crowding of searchers—that occurs whenever “correctness” or “optimality” of a solution is ex-ante unknown.  Furthermore, we find that a critical determinant of crowding is unequal prizes and competition among the agents for the larger prizes. Thus, the phenomenon of crowding is likely to occur whether or not you go through an intermediary, and whether or not you offer single or multiple prizes (as long as there is one largest prize). Our analysis also yields a second robust finding—while breadth of search increases with number of searchers, the relationship is sub-linear (logarithimic) in nature. Thus, looking at only the number of searchers to determine the value of delegated search is likely to over-state the benefit of delegated search-based R&D arrangements.  This study makes several contributions: Firstly, on the theoretical side, we offer a framework of delegated search that accounts for a firm’s choice of extent (or completeness) of problem specification and the agents’ choice of solution approach (i.e., point in solution space to search). Secondly, from the practical perspective of managing the delegated search process, our results offer intuitive rules of thumb on how many awards should be offered, as well as the extent to which firms should undertake problem specification, contingent on the type of problem being broadcast.

Paper 2: Innovation Contests, Open Innovation, and Multi-agent Problem Solving, Terweisch and Xu

In an innovation contest, a firm (the seeker) facing an innovation related problem (e.g. a technical R&D problem) posts this problem to a population of independent agents (the solvers) and then provides an award to the agent that generated the best solution. In this paper, we analyze the interaction between a seeker and a set of solvers. Prior research in Economics suggests that having many solvers work on an innovation problem will lead to a lower equilibrium effort for each solver, which is undesirable from the perspective of the seeker. In contrast, we establish that the seeker can benefit from a larger solver population as he obtains a more diverse set of solutions, which mitigates and sometimes outweighs the effect of the solvers' under-investment in effort. We demonstrate that the inefficiency of the innovation contest resulting from the solvers' under-investment can further be reduced by changing the award structure from a fixed price award to a performance contingent award. Finally, we compare the quality of the solutions and seeker profits with the case of an internal innovation process. This allows us to predict which types of products and which cost structures will be the most likely to benefit from the contest approach to innovation.

Chair: Barrie Nault

Discussant: DJ Wu (Papers 1 and 2 together as one theme)

Paper 1: Open Source Software: Incentives and the Market for Services, Terrence August, Hyoduk Shin, and Tunay Tunca

Abstract: We study the economic incentives of a firm to adopt an open source strategy for software product development. As opposed to the classical proprietary strategy, the main source of revenues in an open source approach is derived from the provision of integration, consulting and support in the software services market. We build an analytical model that captures the fundamental economic trade-offs to the firm in order to explore the economic viability of the open source business model. Our model highlights the negative impact of competition in the services market from third party developers who contribute to the open source project. We also model the economic incentives under a proprietary strategy and demonstrate that a firm can achieve monopoly profits with separate prices for software and services. Our work characterizes the market conditions under which open source and proprietary strategies outperform one another in terms of profitability. Further, our results clarify the critical trade-offs found in an open source strategy, thereby providing firms with guidance on shaping business strategy to support an open source approach.

Paper 2: Lock-in and Software competition: How Does a Proprietary Software Compete against Free Software? Kevin Zhu and Zach Zhou

Abstract: Open source software poses serious challenge to proprietary software vendors. To respond, "lock-in customers and lock-out competitors" seems a viable strategy for proprietary software providers, who attempt to lock in customers by creating substantial switching costs via product design. This paper examines whether such a lock-in strategy will indeed, as commonly believed, benefit proprietary software vendors in the face of competition from not-for-profit open source software. Developing a two-period duopoly model in which customers are heterogeneous in their preferences and willingness-to-pay, we find that the answer could be either way. Lock-in strategy may be counter-productive in competing against open source software. This is in contrast to the competition between two proprietary software providers, whereas lock-in strategy could be beneficial under certain conditions. We find that lock-in reduces social welfare, but certain customers may be better off with lock-in. Finally, we show that the lock-in strategy works differently for different types of customers in the software market (i.e. forward-looking vs. myopic customers). This suggests that customer behavior could significantly alter the equilibrium strategy of software vendors.

Moderator: Vish Krishnan

Ramayya Krishnan, Barrie Nault, Christian Terweisch, Geoffrey Parker

Chair: Kevin Zhu

Paper 1: Core Alignment and Performance in a Tri-Core Model of Reengineering, Victoria Mitchell and Barrie Nault

Discussant: Indranil Bardhan

Abstract: Reengineering projects have proven difficult to complete on time with satisfactory outcomes. Swanson's (1994) tri-core model of IS innovations suggests that a deficiency in one or more of the cores—the administrative core, the technical (business) core and the IS core—can cause process innovations, such as IT-based reengineering projects, to fail. Using reengineering projects in health care, we study the extent to which each core in the tri-core model explains performance, whether alignment among the cores improves that explanation, and how the cores differ in the dimension of project performance they explain. We find that a set of reengineering issues derived from the literature separates into constructs representing the three cores, indicating that know-how to manage these issues are core specific. Next, we find that together the main effects of the tri-core model do not explain project performance, however core alignment in the form of interactions explains both process and product measures of project performance. The process measure of project performance, the duration of project delay, is explained by the IS core and interactions between the IS core and the other two cores. In contrast, the product measure of project performance, client satisfaction, is explained by the administrative core and interactions with the other cores. Thus, the administrative core responsible for management and coordination know-how, rather than the logical IS unit, is responsible for satisfaction with the reengineered process. Finally, we show that the antecedent to the tri-core model, the dual core model where the IS core is embedded in the other cores, does not explain reengineering project performance even when alignment is considered.

Paper 2: IT Investments, R&D Productivity, and Firm Performance, Gokcen Arkali, Indranil Bardhan, and Vish Krishnan

Discussant: Jason Dedrick

Abstract: Prior research has often been based on the assumption that investments in R&D are directly associated with product innovation and revenue growth. While past evidence has supported this view, recent data presents a mixed picture. It is unclear to what extent spending on IT has allowed knowledge workers to become more productive and contribute to the growth in firm profitability. We study the joint impact of IT and R&D investments on firm performance, and investigate whether their interaction makes a difference under different scenarios. We test our model by using archival data from 1998-2004 for a panel of 80 firms across three industries. Our results suggest that IT moderates the relationship between R&D and firm performance in the computers/electronics and pharmaceutical industries in different ways. These results shed new light on the relationship between investments in innovation and firm performance.

Paper 3: An Empirical Analysis of Price, Quality, and Incumbency in Legal Services Procurement Auctions, Vivian Zhong, Tunay Tunca, and DJ Wu

Discussant: Oliver Yao

Abstract: Using online reverse auctions to procure business services, where quality of service plays a major role along with price, is new and emerging in practice. However, the impact of these auctions on the procurement outcomes is unclear. In this context of business services and from the buyer's perspective, we study the impact of online procurement auctions on cost reduction, quality management, and role of incumbent. We find that: (1) On average, hourly rates across all legal service categories are reduced by 13.0% after auctions, assuming a 5% inflation rate for three years; (2) Such cost savings are from both incumbent and non-incumbent firms, with no significant difference between the two groups; (3) Such cost savings are achieved without sacrificing quality; (4) While the overall performance of non-incumbents winner after the auction matches that of before the auctions, surprisingly, overall performance of incumbents winners is significantly improved, suggesting that the value of online auctions comes mainly from incumbent suppliers. Finally, in sharp contrast with well documented incumbent bias on government procurement and online procurement of manufacturing goods, we do not find prior incumbent bias in procuring legal services, and offer novel explanations to incumbent effect.