Organizing Open Innovation: Field Evidence from Hackathons

July 24, 2013 2:07 PM draft

Minjung Choi (minjung.choi@temple.edu) Department of Management Information Systems Fox School of Business, Temple University

Youngjin Yoo (youngjin.yoo@temple.edu) Department of Management Information Systems Fox School of Business, Temple University

Abstract We study the novel open innovation phenomenon hackathon. Hackathon is an open source programming competition that usually lasts 24 hours where contestants pitch ideas and form a team instantly. The hackathon mechanism is an excellent opportunity to look at specifically how digital technology affords a novel form of organizing and how it characterizes the nature of digital innovation. The specific research questions at hand are: (1) what makes it possible to organize innovations in time-pressured hackathon settings? and (2) what is the nature of this open innovation process arising from IT-based real-time interactions? We find two focal technologies that facilitate such a fast-paced mode of development: open web Application Programming Interfaces (APIs) and online source code repository (Gits). To see the interplay between digital artifacts and the emergent organizing process, we will conduct a field study to develop a process model. Three realms of research compose our theoretical underpinnings. First, we look at the IS research dedicated to (open source) software development. We try to fill the gap between the individual level of motivation research and the group level of performance study by exploring how individual designers come to organize to produce innovations, where we believe digital platform serves a key role. Next, we examine the concept of technological affordance and how it transforms the way we organize, work, and innovate. This will enrich our theorization about hackathon organizing by revealing their distinct properties, e.g., a fluid membership, no prior working relationship, very short lifespan, which challenges previous conceptualizations. Last, framing hackathon innovation as distributed knowledge environments, we refer to studies on distributed cognition and knowledge. While prior theories argue knowledge transfer is important, it is also argued to be difficult due to its stickiness. We question whether knowledge transfer is really necessary in such time-critical innovation processes, taking one counter example from the initial observation. To address this paradox, we are drawing upon the modular architecture of software and their programming tasks. Our proposed study will contribute to the IS literature by looking at how digital technology facilitates the emergence of an organization. Going beyond the dated debate over agile methodologies, practitioners can also benefit from our potential insights on self-organized modularization in open innovation processes. We conclude with data collection plans and suggestive directions.

Introduction
April 22, 2013 | This weekend, more than 9,000 people and 484 organizations from around the world came together in 83 cities across 44 countries to engage directly with NASA at the largest hackathon ever held. In just 83 total hours, we collectively tackled the 58 challenges by developing awe-inspiring software, building jaw-dropping hardware, and creating stunning data visualizations that together resulted in one giant leap towards improving life on Earth and life in space. from NASA open government website1 GroupMe was founded by Jared Hecht and Steve Martocci in the Summer of 2010, inspired by a project conceived at the TechCrunch Disrupt Hackathon. After an acquisition in August 2011, GroupMe is now a proud member of the Skype family. from GroupMes company introduction2

Recent modes of innovation are increasingly characterized as open and collaborative (Baldwin and von Hippel 2011), accelerated by digital technology (Yoo et al. 2010). While Apple (2013) proclaims a golden era for a developer community that allows innovators unprecedented new business opportunities, digital innovation is taking diverse forms to elicit creativity and collective power, e.g., online problem-solving contest (Boudreau et al. 2011; Terwiesch and Xu 2008) or crowdsourcing (Archak and Sundararajan 2009; Huang et al. 2012) and with some tweaks as well (Bayus 2013; Park et al. 2013). This open nature of innovation is transforming the way we work and we organize, both challenging and propagating Information Systems research (Information Systems Research Calls for Papers 2013). With this in mind, we study a novel open innovation phenomenon called hackathons. Hackathon, literally meaning a coding marathon, is an open source programming competition for problem solving. The contestants gather in one place and form a team on the spot to create software in a very short amount of time, usually 24 hours overnight.3 They freely pitch ideas at the beginning and aggregate around the attractive ones to work together as a team. Then the newly organized team has to refine the suggested idea, decide upon functionalities and looks of the artifacts being produced, write codes and merge the separate code lines, test usability, finish debugging, demonstrate the prototype; all under extreme time pressure. Google and Facebook have run internal hackathons regularly and sometimes invited outside developers. This tradition has generated such innovations as Facebooks flagship Like button, Timeline interface, Chat feature, and applications for Google Glass (Garling 2011a; Santos 2013). As a novel method of innovation, hackathon is becoming all the more pervasive in other private companies and public sectors, including Microsoft, LinkedIn, Netflix, Yelp, Flickr, LonelyPlanet, cities of New York and San Francisco, NASA, and even the White House.4 It has also widened the applications to music, food, pets, education, fashion,
1

See http://open.nasa.gov/blog/2013/04/22/the-power-of-mass-collaboration for more. National Aeronautics and Space Administration (NASA) just held the worlds biggest hackathon to this date, opening up large stacks of their existing database to encourage civic engagement in addressing global issues. This is one of the efforts recently put forwarded by the federal governments Open Government Initiative (White House 2011; 2012) which reached peak in 2013.
2

See https://groupme.com/about/. GroupMe is a mobile group messaging app and the namesake company. The project won the 1st place at the hackathon run by TechCrunch. Made overnight, it raised US $10.6 M from a number of venture capitalists and angel investors. The next year, the one-year-old startup dramatically grew to be evaluated $80 M when acquired.
3

Conventionally, the connotation of hacking was negative, implying the purported activity to destroy computing infrastructure and do harm to the security. Instead, it has an alternate meaning in this context, i.e., hackers are those who enjoy programming and hacking is their hobby which is sometimes very passionate. For similar use, see (Stewart and Gosain 2006).
4

One source reports that there were 1170 hackathons worldwide in 2012 alone an average of 22 each week (API Evangelist 2012).

health care, etc. to address problems in a given domain. Besides, as carriers of technology venture creation and entrepreneurial opportunities, an example like GroupMe has become a reality. Similar to online innovation contest mechanism (Boudreau et al. 2011; Jeppesen and Lakhani 2010) as it rests on competition among the innovators, hackathons lead to self-organizing of a new group at the onset of a challenge, which questions a previously dominant imagery in organization studies that organization already exists as a static entity. Whereas innovators in the hackathon have to find good ideas as well as qualified team partners instantaneously, this boundary condition provides a good instance to capture the becoming process of an organization in the nascent stage (Weick 1995). Also, as hackathon team members in principle collaborate without much information about each others resource, they may require sensemaking (Weick 1995), establishing task-related knowledge (Wegner 1986), assembling roles (Cohen 2013), and dividing labor (Lee and Berente 2012) very quickly for coordination. IS scholars have recently studied new contexts such as cooperative community (Wang and Ramiller 2009), online social network (Gray et al. 2011), and idea markets (Hellmann and Perotti 2011) in terms of open innovation. Specifically, innovation contests in the digital platform (e.g., TopCoder, InnoCentive, Kaggle) are being a central study subject to attend to the growing industry size (Boudreau et al. 2011; Jeppesen and Lakhani 2010; Yang et al. 2009). In this literature, economic incentives and rewards distribution, competition level, evaluation of idea quality were main foci. On the contrary, we have paid virtually no attention in hackathons given the wide media coverage and calls from practice (Crook 2012). How can a seemingly disorganized group of developers become an organization to create a good piece of functioning software in such a short time? Attempts to account for this question may hint at some of organizations long-held problem statements such as effective individuals vs. ineffective organizations. While the hackathon phenomenon would provoke curiosity from economics, strategy, and software engineering disciplines, we are interested in two specific questions from the technology and innovation standpoint. First, how do hackathon contestant teams organize? Digital technology has afforded novel forms of organizing for innovation (Yoo et al. 2012; Zammuto et al. 2007) and the IS discipline has constantly requested research into new organizing logic in the digital era (Sambamurthy and Zmud 2000). Previous discussions on open innovation organizing have been on the governance in Wikipedia (Forte et al. 2009) and open source community (OMahony and Ferraro 2007; Shah 2006), developers networks (Oh and Jeon 2007; Singh and Phelps 2012), design structure of software products (Baldwin and Clark 2006; MacCormack et al. 2006), and group-level constructs such as effectiveness (Stewart and Gosain 2006) and coordination (Kraut and Streeter 1995), but not on its emergent process and what makes it possible. Exceptionally, Hahn et al. (2008) effectively illustrate the creation of new project teams in open source development finding evidence from prior collaboration ties, which is a stark contrast to our setting where no explicit networks are assumed among developers and designers. By taking a close look at this distinct innovation environment, we believe we can provide a coherent account on the generative mechanism for open innovation and its relationship with the characteristics of digital technology. Second, how does the dynamic and interactive nature of hackathon affect digital innovation? Every open innovation development project has its own style guide and a set of conventions about how to write code, e.g., indentation, annotation, etc. to retain a consistency (Whittaker et al. 2012).5 Given the time restriction, it would be more difficult to coordinate and communicate in the hackathon setting. Hackathon participants are asked to use the cloud computing source code repository such as GitHub on which they can store, import, merge, and distribute codes (Dabbish et al. 2012). Using this technology, the real-time interaction will engender distinct characteristics of innovation processes and outcomes (Leonardi and Barley 2008; Leonardi 2011). To recap, we try to find insights about:
5

Although Google is known for its final products of web applications, the company is also famous for its rigorous method of testing software and code reviews. Code review is a process of peer reviewing codes among software engineers. Normally, peer reviewers create checklists as a review response to improve software (M ntyla and Lassenius 2009). Google provides a code review guideline for developers in its Apps Marketplace webpage. See https://www.google.com/enterprise/marketplace/viewListing?productListingId=5143210+12982233047309328439 &pli=1 for more.

under what processes organizing emerges in a time-critical, competition-pressured open innovation setting, and what is the nature of this digital innovation arising from IT-based real-time interactions.

We conduct an empirically grounded theory building study through a qualitative approach. Van de Ven (1992) suggests it is better to undertake real-time observations of events and activities when processes unfold in their natural field settings. Thus, we adopt field study as the most appropriate methodology to begin the inquiry. Before we start research, we attended a hackathon innovation event sponsored by one of the nations largest technology company and the Philadelphia city authority in April, 2013. We produced field notes from a two-day long participant observation (respectively two hours and eight hours nonstop) and conducted interviews with those selected among participants. In so doing, we not only wanted to describe the occurrence of innovations, but also explain how the materiality of digital technology afforded such characteristics (Gibson 1979; Majchrzak and Markus 2013). This will direct us to theorize about the underlying mechanism that give rise to each event, not only the surface level of observables (Pentland 1999). In what follows, we explain the unique ontological standing of the hackathon phenomenon and the rationale for our empirical investigation. Given the nature of our study, we expect we will iteratively go back and forth from theories to phenomena and vice versa (Lincoln and Guba 1985) as other qualitative studies in the field have done so. Thus, we do not intend to confine our investigation a priori but we briefly present our conceptual background that informed and guided us in conducting of this research. Three realms of research compose our theoretical underpinnings. First, we will see the relevant concepts around knowledge and distributed cognition. This will inform our overall theoretical background and help us investigate how hackathon teams will enact, constitute, and use distributed knowledge in strict time constraints. Next, we will see how the literature on small group has evolved. This will enrich our conceptions on the emergent organization in hackathons and reveal the issues associated with its properties. Last, we will complement our literature review by looking at IS research dedicated to software development. This will be woven together as applicable as we present the initial findings in the subsequent sections. Then, we conclude with implications and contributions along with future data collection plans.

Hackathon as Open Innovation

Hackathons, a longstanding event at the company where hacking is central to the corporate mantra, have been allnight workshops in which employees think up new product concepts and develop rough prototypes. If they impress, those prototypes ended up as commercial products. As the event is a fast-paced contest, Facebook and Google often use it as a recruiting tool to find prospects who fit with the breakneck pace of their corporate culture (Garling 2011b). In an in-depth ethnographic study on Facebook, Fattal (2012) concludes the word hack points to this companys internal cultural logic, reproducing the founding ethos as well as its management approach to seeking for better solutions (Facebook 2012, p. 67).6 The concept of hackathon is in fact not completely new to developers community. They have organized independent coding fests, where they can boast and play with their programming skill. This has comprised programmers subculture (Thomas 2003). Research at hacker conferences highlights how face-to-face encounters and collaboration in limited physical space help bond hacker communities often divided by geographic distance (Coleman 2010). This is consistent with a large body of IS literature on software development teams (Crowston et al. 2007) and Facebooks hackathon serves a similar purpose letting coworkers collaborate
6

From our empirical observations, we identified the colloquial usage of the word hack was different in the hackathon participants discourse from that of the public at large. A hack can refer to a chunk of coding job or its successful outcome as a solution to a programming problem. Technologists and developer community tend to endorse their own set of jargons which will not be easily acceptable by a larger community, and this also contributed to forming a social identity as hackers. As field researchers, we h ad to become familiar with those terms, not only to immerse ourselves in the context but also to truly interact with our informants (Klein and Myers 1999). Our interview transcripts also indicated that they do use these jargons frequently without much cognitive attention. We will introduce some examples later on.

beyond their small teams tightly defined objectives (Fattal 2012). Also, as a ritualized side project, hackathon sparks excitement among employees, potentially reducing the risk of job burnout (Maslach et al. 2001) common in IT professions. Although anthropological investigations into hackers depict they hold a maniacal sprint of coding (Coleman and Golub 2008, emphasis added), hackathons are being considered a dependable way to effectively and comparably quickly solve organizations R&D problems as well as unsolved civic issues. In addition to entailing inviting multiple actors beyond the organizational boundary, there is another reason to refer to hackathons as an open innovation phenomenon from a technological point of view. First, the new programming philosophy based on open web APIs (Application Programming Interfaces) is completely changing the way engineers work on innovation.7 What used to be very complicated can be done simply by one line now because open APIs are freely shared, can be flexibly edited and improved, and applied to generate unlimited uses. With the technological advancements with APIs, software development has become a game of mix-and-match, with a creative twist. What matters is not only how good you are in coding, but how much APIs you know and how to generatively reprogram them for localized usages. The reuse of codes through APIs is a form of knowledge reuse in software development fundamental to innovation (Haefliger et al. 2008), and is a very important technological element that enabled fast-paced hackathon innovations. The second technology artifacts that are used for hackathon innovation is Gits. This is a repository of codes developed in previous hackathons. Most apps developed in hackathons are based on previously written codes. As the programmed codes provide a common ground for communication among engineers digitally, coordination patterns will be different from the contexts that were previously studied although they are in the same space and time. Traditionally, software development was deemed to be monitored and controlled by management (Choudhury and Sabherwal 2003; Gopal and Gosain 2010; Harris et al. 2009; Ji et al. 2005; Maruping et al. 2009). Abundant research on development methodologies evinces this view that effective methodologies do exist and can be identified (Fitzgerald et al. 2006; Ramesh et al. 2011). This research trend gave birth to the debate between plandriven and agile methods (Agile Manifesto 2001), and more commonly on the appropriateness and effectiveness of various agile methods and practices (Vidgen and Wang 2009). This stream has been particularly informative in assessing the team performance at a group level (Faraj and Sproull 2000; Guinan et al. 1998; Kraut and Streeter 1995). Similarly, IS literature on open source development has been majorly delineated by why open source developers participate in this voluntary activity (Aksulu and Wade 2010), which primarily looked into the individual level of motivation. Although these separate lines of efforts are valid in their own right, IS field is suffering a myopic view instead of demonstrating a holistic picture, without knowing in detail how individual designers get organized to interactively produce innovation in a team. Also, design environments are dramatically changing with multiple actors engaged in distributed development (see Information Systems Research Calls for Papers in 2006 and the journals Special Issue on Flexible and Distributed Information Systems Development in 2009) affecting its diffusion and evolution as well (Boland et al. 2007; Yoo et al. 2012). Consequently, static views on the organization can be limited in explaining complex phenomena that are actually embedded and bounded by contextual factors such as a temporal condition (Orlikowski and Yates 2002). Therefore, we intend to turn our attention to the procedural analysis of innovation, from the emergence of an organization to its evolvement into a devoted unit constructing innovations on a strict timeline. This led us to the second part of our theoretical underpinning.

Hackathon as New Organizing

If theres a better way to transition from a group of strangers to a tight-knit group of friends in 72 hours, I have yet to see it.
7

API is a chunk of codes that composes the interface part of software applications. This can be understood as a modular element of what developers would like to present their programming to users. Because it is communicable with each other, the production of APIs means you can generate virtually unlimited variations. Google Maps APIs are a good example, which have been embedded in countless websites to be used for different applications.

 a serial hackathon participant and entrepreneur 8

Both hackathons and the emergent teams sketch a very unique organizational context, which complicates our theoretic framing. Its temporary nature touches upon project-based organizations (Lundin and S derholm 1995) and its observed attributes implicate the overlap with small group development (Tuckman and Jensen 1977). While these conceptions still hold to some extent, its internal dynamics do not resemble them as much as we wish, owing to its distinctive organizing logic facilitated by both fast-paced contexts and its close connection to digital technology. A hackathon normally starts with the presentation of its purpose, requirements, judgment criteria, and prize arrangements. Follows the idea pitch for several minutes in which participants freely throw out innovation concepts. They then freely form a group according to estimated expertise or preference about the ideas. The newly created teams work together throughout the challenge to produce mobile apps, websites, APIs, etc. Evaluation of submitted works is based on how much they used the suggested APIs. Equally important are completion and demonstratability of the program. Because it happens only within a short period of time, sometimes you cannot meet the deadline. Thus, you need to know how much you will go further and sometimes you ought to compromise functionality and given time depending on available capacity of your team. Although every team is geared toward winning the competition, only those selected will be declared winner or awarded, so the evaluation and decision about team partners and suggested ideas become highly important. However, with its instant and simultaneous characteristic, the organizing process may take place without well-structured choice, coupled with the lack of information about each others expertise (Faraj et al. 2011). Consequently, the resultant organization may need to quickly make sense of what is going on (Weick 1995), identify task-related knowledge (Wegner 1985), or flexibly improvise to a certain extent (Weick 1998). Participants may depend on intuitive hunches (Locke et al. 2008) or subjective assessments (Elsbach and Kramer 2003) rather than rational decision making. Given the time constraint, judgments must be made before actual products are produced or reliable information is available. Ironically, this uncertainty is a key element in mobilizing the emergence of an organization in hackathons. In fact, sociologists have documented on how small groups naturally take shape. On the one hand, group formation comes out of bonds of interpersonal attraction. On the other hand, a group starts when individuals perceive that they share some social category such as nurses or baseball players (Hogg and Williams 2000). Regardless of either view, a dominant notion is that group formation is preceded by the identification of oneself and others. Management and organization theory has also explored the forces behind group formation, with a little tweak toward the managerial setting. Many theorists argue that social contract is what makes people consent to join and remain in organizations. For example, Etzioni (1964) posited organizing is done in order to promote goal attainment. In reviewing such divergent views, Weick (1995) concludes that there will be a continuing tension between perceiving similarities and dissimilarities among members in a small group. Although all of them are adequate illustrations, integration of inconsistent views will yield illogical reasoning since underlying assumptions and orientations significantly vary. Hence, we are cautious about integrating and applying extant theories crudely given the risk of losing diverse angles each hints at. While we recognize the relevance of prior recounts, we want to discover a distinct mode of organizing that hackathon-based open innovation is based upon. The emergent organization coalesces and disbands very quickly and exits only shortly, and this boundary condition will weaken and modify prevailing assumptions held in mainstream organization theory, e.g., structural roles, coordination, knowledge transfer, to name a few. There is another reason we are being careful. A common assertion in literature about groups (e.g., Lorge et al. 1958) is that most researchers study ad hoc, short-lived, temporary groups because these groups behave differently from groups of longer duration. Duration of contact may be a more crucial determinant of group behavior than the number of people (Hall and William 1966; Gartner and Iverson 1967). Our point is not that size is irrelevant but that while there are many ways it may have distinct effects, size per se is a misleading variable if it is used as the starting point of an inquiry.
8

Read at http://venturebeat.com/2013/03/30/entrepreneurs-stop-participating-in-hackathons-just-to-win-them/ for a more lively description.

It is probably more important to note that this impromptu organizing reflects upon the new technology arrangements that are completely unprecedented (Yoo et al. 2010; Zammuto et al. 2007). Online communities rest on distributed organizing without a fixed membership (Faraj et al. 2011), massively multiplayer online role-playing games (MMORPG) pursue task-based ad-hoc virtual teams (Huang et al. 2013), and collaborative digital artifacts enable us to envision emergency response groups to unexpected requests (Majchrzak et al. 2007; Faraj and Xiao 2006). This emergent group arises from a spontaneous process of group formation and is characterized as having no preexisting structure or prior experience of working together. Yet, these groups still express high levels of interdependence and coordinate knowledge, resources, and tasks (Majchrzak et al. 2007). How is this possible? One explanation is through the lens of technology affordances. Originally put forward by Gibson (1979), technology scholars have adapted and developed the concept to explicate how the materiality of information systems, technological artifacts, and digital objects enact organizations specific uses. As IS scholars, our focus is not only on the organizing per se but also on how technology afforded its consequences (Majchrzak and Markus 2013). By organizing, we refer to the generative mechanism under which organizations emerge and are constructed to be able to innovate under high pressure. What is the distinctive manner by which this type of organization operates? In what way will the inner workings of this organization affect engendering innovation in such a competitive environment? Taking this into account, we are curious to know how different individual minds in this dynamic setting come to create innovation collectively (and successfully), which connects us to the last but most important pillar of our theoretical background.

Hackathon as Distributed Knowledge Environment

According to Nye and Brower (1996), an ideal setting for studying the social and interactional aspects of social cognition is the small group. Regarding the cognition distributed among multiple actors within a hackathon innovation team, anthropological studies of Hutchins (1991; 1995) informed our perspective on knowledge and cognition. In these accounts, knowledge is distributed and cognition is socially and culturally constituted through ongoing everyday practice. This practice view has been reflected in the IS and knowledge management research, for instance in Orlikowski (2000; 2002; 2007), Levina and Vaast (2005), and Carlile (2002). We intend to use this lens to understand how the hackathon innovators enact knowledge as they dynamically engage in real-time interaction both using and creating technological artifacts. Argote and Miron-Spektor (2011) establish that knowledge could be embedded in a variety of repositories, including individuals, routines, and transactive memory systems. Transactive memory is a meta-knowledge on who has better expertise on what, which is believed to increase efficiency and effectiveness of a task group (Wegner 1985). Previous studies explored the antecedents of TM such as prior experience (Liang et al. 1995) and cognition-based trust (Kanawattanachai and Yoo 2007), presumably to be applied in later tasks. Problematic is that we often do not know what we know (Sch n 1983) or even what we do not know (Rumsfeld 2003) until we systematically verbalize it. Polanyi ([1967] 2009) makes a similar argument pointing to our ability to ride a bicycle even when we cannot articulate how we do it precisely. In lieu of this, a good amount of Weicks research asks a question he has labeled the sensemaking recipe: How can I know what I think until I see what I say? (e.g., Weick 1979; 1995). This highlights how sensemaking is a retrospective process in which individuals first act and then reflect on their actions to interpret what they mean. However, for the TM or situational sensemaking to be recognized, organizations should not only learn who knows what but also decide who will do what (Jarvenpaa and Majchrzak 2008; Majchrzak et al. 2007). This necessitates not only the prior existence of TMS but also the decision on the division of cognitive labor required for innovation (Lee and Berente 2012). Extant studies conceive of a step-by-step model in which knowledge is readily built first and used later. In this world, there are separate, sequential stages between establishing and applying TMS to doing the task.9 This does not address the fluid nature of knowledge coordination (Faraj et al. 2011) in a more simultaneous setting.

See Brand and Hollingshead (2004) for a detailed discussion.

We wish to address the shortcomings of the normative models by stressing more on the behavioral aspects of knowledge. For example, the iterative, cyclical imagery of time dimension would be inappropriate in a highly temporary organization like hackathon project teams as the organizing occurs one-time only. Rather, we propose that TMS is enacted simultaneously through a series of practices of action. This echoes Weick and Roberts view that actions can be knowledge (1993). Yoo et al. (2012) bring forth the refreshing notion of a team in light of distributed innovation. They propose an innovation team can be reconceptualized as a highly varying task-expertisepeople unit (Brandon and Hollingshead 2004) which is fluidly combined, carried out, and replaced in some emergent trajectory toward a goal or problem resolution. According to this assertion, each hack produced in hackathons constitute an innovation trajectory. In this trajectory of innovation, innovators embed their tacit knowledge. By doing so, organizing is constituted in the practices of programming and coding activities (Orlikowski 2002). Knowledge as well as transactive memory emerge from the ongoing and situated actions of organizational members as they engage the world. This will be our vintage point in studying the organizing in hackathon innovation teams. Our specific focus will be on how meta-knowledge on each others expertise is enacted through practices of minute levels of tasks, and in such processes how digital technology (code sharing online repository) would facilitate it.

Methods and Empirical Context

Methodology
We view it essential to adopt a qualitative and inductive approach to our inquiry. In particular, we argue that previous, laboratory studies have been inadequate to examine two important features of organizing that come about in real-world hackathon settings. First, interpersonal judgments of hackathon participants expertise typically involve dynamic processes that are strongly dependent on context (that is, they are subject to localized and situated norms and expectations). Second, these organizing processes unfold over time and involve extensive interaction among engaged actors. Traditional survey research designs and statistical analyses are static and therefore not readily suited to examining such dynamic and evolving phenomena. In contrast, qualitative research designs have been particularly well suited to analyzing dynamic, interactive processes (Lee 1994; 1999). By this reason, we chose qualitative logic to see organizing and creating innovation. For data collection, we will employ participant observation. Semi-structured interviews (Myers and Newman 2007) will follow if necessary, with the participants prior consent to contribute to the research. We have prepared interview protocols informed by our initial literature review. Yet, as we go on to a site visit, we intend to include new concepts that we found interesting conducting additional search for theory in parallel.

Site
We attended a hackathon innovation sponsored by one of the nations largest technology company and a major funding agency with the citys authority to get a flavor about what a hackathon is and devise research plans. The event consisted of a two-day session, one night for idea pitch and networking and the other eight-hour for development and presentation. We counted the number of participants regularly so that we can gauge the turnouts and possible dropouts. About 60 people remained in the next day while 80 people stayed on average during the previous nights pitch. This is in part attributable to the event was rather open and informal although registration was recommended. A total of 12 teams were formed and competed finally, each of which had varying numbers of members ranging from two to six. Below is the sequence of events and issues to explore. Activities can occur concurrently, iteratively, in a reverse order, or do not happen at all. The issues are suggestive so still remain to be refined and we do not necessarily intend to grapple with all listed issues separately in our subsequent research.

Table 1 Hackathon Organizing Timeline and Suggestive Issues Timeline Idea Pitch (1 hour) Activities identified Idea generation Idea sales Team partner recruiting & selection Team formation (1 hour) Idea refinement Decision on interface and functionalities Exemplar issues * How does one assess another persons creative potential and compatibility? * How does one dynamically engage in the organizing process? * How the roles and tasks are modularly identified, enacted, substituted, merged around the focal technological artifact (Gits)? * How do innovators with expertise situate their tacit knowledge in the design process, especially in the planning and gauging the feasibility? How does this affect the completion (and success) of innovation? * How ideas survive and die out throughout the whole processes given technological constraints? Development (5-6 hours) Interface design Coding Prototype building Bug Fixing (2-3 hours) Codes migration Usability test Debugging Negotiation * How is knowledge enacted through activities of designing and coding? * How is transactive memory constituted through real-time interaction using digital technology as boundary objects? * How are ideas negotiated, especially at the last minute debugging stage before submission? * How does coordination occur in this highly instant context?

Result (1 hour)

Presentation Demonstration Evaluation

* How is behavioral knowledge transferred, if ever?

Initial Observation and Interviews

As a field researcher, we took into account the interaction between the researcher and the subjects (Klein and Myers 1999). We recognized from the research design that we should be able to understand the contextual nuance embedded in hackathon innovation, as most of the contestants were software engineers who are likely to endorse a certain norm (Stewart and Gosain 2006). We attempted to learn not only their jargons and technical terminologies but also their sentiment by joining online hacker communities and browsing popular hacker blogs. Also, we tried to immerse ourselves in their culture by referring to historic accounts on hackers and programmers (Levy 1984; Thomas 2003) before we actually entered into the site. As a result, we encountered such terms as spaghetti code, which did appear in one of our interview transcripts.

On site, we engaged in informal conversations with the hackathon participants. On the next day, we landed on two different teams to make observation, producing field notes occasionally. After the event, we obtained contacts of selected participants for subsequent interviews. Also, we gathered information on the web documenting the event, before and after, and complemented our real-time observation via photography and videography for archiving purpose. 1985). Our interviews yielded about 100 pages of single-spaced pages of transcript data. After the initial data collection, we had to revisit the literature to modify and sharpen our angles. For instance, one of our intentions was to see how roles or tasks are flexibly allocated and possibly overlapped around the programming to meet the time restriction. We had then looked at the traditional small group theories, but we found such a stream would not be adequate to elucidate hackathon organizing in particular. The dominant imagery in this arena is that a manager places well-understood tasks into stable roles, which has produced a rather prescriptive account (e.g., Hackman and Oldham 1975; Williamson 1975). Even much of the current theory still holds on to the conceptualization of jobs and tasks as the product of deliberate, goal-directed managerial action (Cohen 2013). There is ample evidence, however, even in very similar circumstances, organizations use the same technology to perform the same task but structure that work differently (e.g., Barley 1986; Orlikowski and Robey 1991). The hackathon organizing was indeed highly chaotic, for example, one participant failing to partner with anyone, one team still looking for a member as one person did not show up as promised, and another team not knowing what to produce until the official beginning. Roles and tasks were not even discussed in one team whereas another team was quick to identify, showing different paces and patterns. The realization from both the theory and messy reality that organizing is situated and constructed, and is not about being but becoming (Weick 1995) let us strengthen our perspective as a constructivist. We prepared interview protocols with the help of relevant studies attaching protocols (e.g., Crowston et al. 2007) as well as our fields epistemological guidelines on interpretive research (Klein and Myers 1999; Myers 2009). All interviews were recorded and transcribed, and kept confidential. Interviewees were asked to recall and reconstruct the timeline from the beginning to end, with interventions by the interviewer if necessary. Table 2 is an overview of the hackathon project team profile we selected (all under pseudonyms). Table 2 Subjects Profile Subject Dan T. Kai M. Ian Z. Harry R. Description Software designer Business person Software designer Hardware designer Interview April 8, 2013 April 12, 2013 electronically communicated electronically communicated Duration 75 minutes 60 minutes N/A N/A

Discussion
In this section, we discuss how hackathon organizing makes distinction from the conventional sense and what rising issues we encountered.

Knowledge Sharing and Constituting Transactive Memory

One informant mentioned the situation when he first met the third person in his team: Ivan was good at coding and he had met Kareem once, and I liked him. When asked how he knew about the third persons expertise, he described in detail about his pair coding experience during the hackathon. As they worked on the same repository, sharing each generated file and checking real-time improvements, he realized that the third person was logical, meaning the code lines do not have a conflict and would function without error. Here, without actually running the program, he could almost intuitively know that the code was well written because he also had the expertise to in

10

programming. This is contrasted with what we heard from the interview with the first person, who was not knowledgeable at computer coding. Kareem, as a non-designer, did know Ivan was good at coding when asked retrospectively he could not explain in detail but just stated that he had heard from other members. Dave, a seasoned engineer, on the other hand, not only knew Ivan was good at coding but also to what degree he was good at that job. The sophistication levels of transactive memory constructed from a designer and a non-designer differ because of their difference in expertise of design. We could see the quality of transactive memory was dependent upon both the individuals applicable expertise and his additional ability to accurately evaluate that expertise, which was highly situated and sometimes hard to be verbalized. Here, transactive memory was embedded in the material artifact they used (code lines and digital platform). By improving the quality of codes, Dave could enact knowledge about how to write codes better looking at the real-time code updates, constitutively updated the understanding of another person s expertise accordingly, and the system they were building could be improved at the same time. This coevolution of organizational knowledge, transactive memory, and the system is indicative of sociomaterial coevolution around the focal technology (Orlikowski and Scott 2008).

Modularity of Innovation Labor and Coordination

A distinctive feature of this novel form of organizing is that the innovators are physically collocated but work on the completely different parts in the development (e.g., software and hardware, front-end development and back-end development). For instance, in the case of pair coding, which is the most common case, they are present at the same place and time but they work on the codes separately in an independent technological setting (but still connected and the outcomes synchronously shared via the Internet). On the contrary to our initial belief that the innovators will need to identify task-related knowledge to divide roles and tasks due to the lack of previous experience together, they implicitly knew where to contribute by self-selecting task division. This is essentially possible because of the architecture of the digital artifact they create computer programming can be modular. Modularity points to individual components that make up a larger whole and the capability to divide tasks and labor accordingly as each holds interdependence. Drawing upon the architecture of complex systems (Simon 1962), modularity has been heralded as an organizational and technical architecture that enhances incremental and modular innovation (Ethiraj et al. 2008). Modularity enables self-selection, allowing for greater fit between a task and a developers interests skills. To understand implications of modular task architectures is essentially important in open source systems development because developers (thus, heterogeneous knowledge resources) are distributed (Yoo et al. 2008). The code architectures of products produced through open source model is often claimed to be more modular than those written within organizations (OReilly 1999; Raymond 2001). MacCormack et al. (2006) explains this is because, the sharing of information about solutions adopted in different parts of the design is much easier and may be encouraged due to the physical proximity in proprietary 10 development model. In lieu of this, research on open source development emphasizes creating common grounds (i.e., knowledge sharing) among individuals to enable coordination (Hinds and Kiesler, 2002; Srikanth and Puranam 2011) with a stress on the lack of face-to-face interaction. Consequently, in many of these accounts, importantly studied is the role of collaborative technologies such as email, bulletin boards, and version control software (Lee and Cole 2003; Raymond 1999; Shah 2006). Consistent with this view, media synchronicity theory (Dennis et al. 2008) argues that the more synchronous communication channel will create richer messages. The hackathon innovation creates an interesting tension here, as it is a distributed work environment where tasks are highly modular but knowledge sharing through face-to-face interaction is readily possible and thought be necessary to coordinate. However, on the contrary to the previous theories, there was little knowledge being shared about the overall system being developed although there was a
10

In this research we are referring to open as the mode of organizing as Chesbrough (2003; 2005) did but prior research has tended to entail the provision of the product as well, meaning open access to users and shared for free. Thus, open source innovation is frequently used as opposed to proprietary software although open development can charge a product and intra-organizational development can distribute free copies. In this context, proprietary mode of development points to the dedicated teams of developers.

11

highly synchronous channel available. Innovators are clearly aware of their tasks and their possible contributions although they did not explicitly set a goal, do not have trust derived from prior experience as behavioral and psychological scholars may argue. Of course, the availability of electronic communication technologies is no guarantee that knowledge sharing will actually take place (Alavi and Leidner 1999; Orlikowski 1996). To elucidate this, existing knowledge management research has supported and reinforced the belief that individual motivation matters (Wasco and Faraj 2005). Surprisingly, however, without much discursive communication as well as shared future plans and previous commitment, the hackathon innovators were able to collectively accomplish the production of a complex artifact (the software) very successfully (for affirming evidence, see Bolici et al. (2009) and Bruns (2013)). Looking at the behavior of a group level, they seem to be cooperating in an organized and coordinated way; yet at an individual level, they work alone as if they have an independent task. This contrast between the individual and the collective level has been studied for a long time in biology where it has been called the coordination paradox or stigmergy (Grass1959; Theraulaz and Bonabeau 1999). We argue that this was particularly made possible due to both the characteristics of the product they produce and the method they use, i.e., digital technology. It is noteworthy that this pattern change towards the end of the competition as they had to merge the separately produced codes, thus resulting in integrating individual tasks accordingly. In the last-minute debugging processes, there was dramatically high interaction and negotiation over the codes and features to meet the timeline. In this stage, cooperation was constituted by the increased interdependence of the designers who changed the state of their previously modular tasks, and interact on a common field of work (Schmidt and Simone in 1996, p. 158) on the repository.

Implications and Limitations

The hackathon phenomenon is as an excellent opportunity to look at specifically how digital technology affords a novel form of organizing and how it characterizes the nature of innovations. It begets a new innovation context bounding different conditions to organize, particularly temporally. The actors show different patterns even to achieve the same goal than do traditional organizations, where we find opportunities for theory development. We contribute to theory by looking at what transpires in this new organizing context for open innovation. As Dougherty (2006) puts, organizing for innovation should be defined in its own terms, not in terms of the deviation from the conventional model. We study when and under what conditions innovation would emerge, facilitated by two focal technologies, and this adds to the IS literature. From a practice view (Feldman and Orlikowski 2011; Orlikowski 2000) based upon practitioners field activities, we advance the managerial understanding on under what conditions uncoordinated individuals to become coordinated. Especially, we believe potential insights can be transferrable to existing sorts of organizations as well, e.g., newly created technology venture firms. Entrepreneurial organizations in the nascent stage can benefit from our study as we explore how hackathon innovators enact ability to coordinate in a very instantly organized setting. Technology ventures should survive in time-critical environments under intense competition, meaning not only the product development but presentation to markets is important. Hackathon innovation challenges let participants develop IT artifacts very quickly. They should make sure that the prototype does not function erroneously, and if ever, respond and discuss quickly within the team. It echoes the latest efforts made by startup incubators and venture capitalists. Y-Combinator, AngelHack Accelerator, and StartupBus are sponsoring hackathons to locate funding targets as well as to test the team members organizing capability to turn the new ideas into decent software. It is important to recognize the potential limitations of our study which will further guide our next rounds of data collection. First, given the nature of interpretive research, the study will be strongly data-driven until the investigators feel the data reached saturation (Glaser and Strauss 1967). Theoretical saturation is achieved is best described as the point at which diminishing returns are obtained from new data analysis, or refinement of coding

12

categories. Gasson (2004) maintained that the point of diminishing returns comes when (and only when) theoretical constructs fit with existing data and the comparison of theoretical constructs with new data yields no significant new insights. If a researcher fails to look in the right place, however, nothing may be recorded and if the conceptual category is too broad, it can include confounding data (Van de Ven 1992). Also, as we depend on multiple informants to reconstruct the whole process of organizing and enacting knowledge, practical tradeoffs such as cost effectiveness can stand in the way (Pentland 1999). Second, our preliminary findings suffer from the absence of archival evidence. In the next site visits, we also plan to collect textual data stored on the code repository so we can trace revision versions and times in their innovation trajectory (Yoo et al. 2008; 2012) and perhaps match our field observation data.

Conclusion
Data Collection and Analysis Plan
We are planning to get authorization to participate in additional hackathon events. Table 3 is a selected list of hackathon events planned in 2013 near Philadelphia and major cities in the United States. Before we actually go into sites, we will prepare field protocols guided by our initial findings and the rising concepts we encountered. Our interviews will also feature more extensive literature review as we narrow down a few concepts.

Table 3 Upcoming Hackathon Competitions in Philadelphia, New York, Boston, DC, and Silicon Valley Date April 5-7 April 20-21 April 26-28 April 27 May 18-19 June 1-2 August TBA September 6-8 October TBA November TBA Dec 7-8 June 1-2 June 1-2 June 1-2 June 15-16 June 22 Theme AT&T EduTech Hackathon NASA Space Apps Startup Weekend News Hackathon Music Hack Day Random Hacks of Kindness Apps for SEPTA PennApps Fall Hacks for Democracy Startup Weekend Random Hacks of Kindness AngelHack NYC AngelHack Boston Hack for Change DC AngelHack Silicon Valley Windows 8 App Factor Organizer(s) AT&T, Technically Philly, Jarv.us Azavea various volunteers Technically Philly ExCITe Center at Drexel University Technically Philly (National Day of Civic Hacking) Jarv.us University of Pennsylvania student group Azavea various volunteers Technically Philly AngelHack AngelHack White House (National Day of Civic Hacking) AngelHack Microsoft

13

June 21-22

Intel Perceptual Computing Workshops and HackNight

Helios Interactive (San Francisco), Intel (Santa Clara)

In parallel, the current round of literature review will continue. To rigorously verify findings as well as discover insights that will arise as data accumulate, we will apply line-by-line coding to interview transcripts, observation field notes, meeting reports, and written archival material. For analysis, all field notes and interview transcription will be uploaded into the latest version of Atlas.ti, a widely used computer-assisted qualitative data analysis tool. The software has a trial version with limited functionality and it lets us analyze textual data by visualizing them in a more structured way. Figure 1 displays sample interface of the program.11

Figure 1 Sample Interface of Atlas.ti There are a multitude of strategies to analyze qualitative data depending on purposes. Mostly, interviews are read thoroughly several times, and the coder identifies and lists recurring themes in the data using a grounded-theory categorizing process (Strauss and Corbin 1998). Sometimes scholars pursue triangulation (Denzin [1978] 2006) to

11

Note that there is no formal notation for interpretive coding. Therefore, the directions in the bottom window do not indicate a hypothesis-based causal relationship as statistical path modeling does. Also, the small text boxes are not equivalent of measured variables in a positivist sense but represent different categories and social constructs that the researcher find interesting.

14

provide richer explanations complementing what quantitative methods cannot tap into. Slaughter and Kirch (2006) quantitized data by converting their collected qualitative information into numerical codes and numbers that could be statistically analyzed. Some research use interview data for further exploratory factor analysis (EFA) by developing new constructs (e.g., Arnold and Renolds 2003). All into consideration, we seek to collect version history and revision records archived in the open codes repository used in each hackathon challenge. Our initial site visit was informative of how they coordinate without preconceived coordination but we could not look more closely into how technology came into play their activities. Thus, we would like to see how digital technologies facilitate organizing using the data obtained from open access code repository and explicate their innovation process.

Next Directions and Expected Outcome

We aim to develop a process model to toy with the interplay of digital technology and organizing activities. What makes it easier? What makes it more difficult? What are the antecedents of this organizing? What makes hackathon innovations successful? Do they feature distinct characteristics? If yes, in what dimension? Still broad, these are our next questions in mind. We will collect data until we believe we would be confident about our propositions with more established evidence. As we go along the way, we will be sensitive to any new findings that will arise, which did in our initial results. As we have briefly seen above, we find interesting the coevolution of systems and knowledge concurrently, made possible by digital technology which we see the connection with sociomateriality. Also, the generativity (Zittrain 2006) of APIs and Gits used in hackathon innovations, so we will delve into how hackathon participants use and communicate with these technologies more in detail. Thus, we expect next drafts will include extensive literature review on, but are not limited to, hacker culture, sociomateriality, modularization of innovation.

15

References
gerfalk, P. J., Fitzgerald, B., Slaughter, S. A. 2009. Introduction to the Special Issue: Flexible and Distributed Information Systems Development: State of the Art and Research Challenges, Information Systems Research (20:3), pp. 317328. Agile Manifesto. 2001. Manifesto for Agile Software Development, http://agilemanifesto.org/, retrieved May 12, 2013. Aksulu, A., and Wade, M. 2009. A Comprehensive Review and Synthesis of Open Source Research, Journal of the Association for Information Systems (11), pp. 576-656. Apple. 2013. App Store Tops 40 Billion Downloads with Almost Half in 2012, http://www.apple.com/pr/library/2013/01/07App-Store-Tops-40-Billion-Downloads-with-Almost-Half-in2012.html, retrieved April 26, 2013. Archak, N., and Sundararajan, A. 2009. Optimal Design of Crowdsourcing Contests, in Proceedings of the 30th International Conference on Information Systems (ICIS), Phoenix, AZ. Argote, L., and Miron-Spektor, E. 2011. Organizational Learning: From Experience to Knowledge, Organization Science (22:5), pp. 11231137. Baldwin, C. Y., and Clark, K. B. 2006. The Architecture of Participation: Does Code Architecture Mitigate Free Riding in the Open Source Development Model?, Management Science (52:7), pp. 11161127. Baldwin, C. Y., and von Hippel, E. 2011. Modeling a Paradigm Shift: From Producer Innovation to User and Open Collaborative Innovation, Organization Science (22:6), pp. 1399-1417. Barley, S. R. 1986. Technology as an Occasion for Structuring: Evidence from Observations of CT Scanners and the Social Order of Radiology Departments, Administrative Science Quarterly (31) pp. 78-108. Bayus, B.L. 2013. Crowdsourcing New Product Ideas over Time: An Analysis of the Dell IdeaStorm Community, Management Science (59:1), pp. 226244. Boland, R. J., Lyytinen, K., and Yoo, Y. 2007. Wakes of Innovation in Project Networks: The Case of Digital 3-D Representations in Architecture, Engineering, and Construction , Organization Science (18:4), pp. 631647. Bolici, F., Howison, J., and Crowston, K. 2009. Coordination without Discussion? Socio-technical Congruence and Stigmergy in Free and Open Source Software Projects, in Proceedings of the 31st International Conference on Software Engineering (2nd International Workshop on Sociotechnical Congruence), Vancouver, Canada. Boudreau, J. K., Lacetera, N., and Lakhani, K.R. 2011. Incentives and Problem Uncertainty in Innovation Contests: An Empirical Analysis, Management Science (57:5), pp. 843-863. Brandon, B. P., and Hillingshead, A. B. 2004. Transactive Memory Systems in Organizations: Matching Tasks, Expertise, and People, Organization Science (15:6), pp. 633-644. Bruns, H. C. 2013. Working Alone Together: Coordination in Collaboration across Domains of Expertise, Academy of Management Journal (56:1), pp. 62-83. Choudhury, V., and Sabherwal, R. 2003. Portfolios of Control in Outsourced Software Development Projects, Information Systems Research (14:3), pp. 291314. Cohen, L. E. 2013. Assembling Jobs: A Model of How Tasks Are Bundled Into and Across Jobs, Organization Science (24:2), pp. 432454. Coleman, G. 2010. The Hacker Conference: A Ritual Condensation and Celebration of a Lifeworld, Anthropological Quarterly (83:1), pp. 4772. Coleman, E. G., and Golub, A. 2008. Hacker Practice: Moral Genres and the Cultural Articulation of Liberalism, Anthropological Theory (8:3), pp. 255-277. Crook, J. 2012. From Disrupt NY To A Multi -Million Skype Acquisition, GroupMe Tells All, TechCrunch, http://techcrunch.com/2012/05/11/from-disrupt-ny-to-a-43-million-skype-acquisition-groupme-tells-all/, retrieved March 23, 2013.

16

Crowston, K., Howison, J., Masango, C., and Eseryel, U. Y. 2007. The Role of Face-to-Face Meetings in Technology-Supported Self-Organizing Distributed Teams, IEEE Transactions on Professional Communication (50:3), pp. 185-203. Dabbish, L., Stuart, C., Tsay, J., and Herbsleb, J. 2012. Social Coding in GitHub: Transparency and Collaboration in an Open Software Repository, in Proceedings of the 15th ACM Conference on Computer Supported Cooperative Work (CSCW), Seattle, WA, USA. Elsbach, K. D., and Kramer, R. M. 2003. Assessing Creativity in Hollywood Pitch Meetings: Evidence for a Dual Process Model of Creativity Judgments, Academy of Management Journal (46:3), pp. 283-301. Facebook. 2012. Registration Statement on Form S-1, United States Securities and Exchange Commission: Washington, D.C., USA, available at http://www.sec.gov/Archives/edgar/data/1326801/000119312512034517/d287954ds1.htm, retrieved May 12, 2013. Faraj, S., Jarvenpaa, S.L., and Majchrzak, A. 2011. Knowledge Collaboration in Online Communities, Organization Science 22(5), pp. 12241239. Faraj, S., and Sproull, L. 2000. Coordinating Expertise in Software Development Teams, Management Science (46), pp. 1554-1568. Faraj, S., and Xiao, Y. 2006. Coordination in Fast-Response Organizations, Management Science (52:8), pp. 11551169. Fattal, A. 2012. Facebook: Corporate Hackers, a Billion Users, and the Geo-politics of the Social Graph, Anthropological Quarterly (85:3), pp. 927956. Fitzgerald, B., Slaughter, S. A., and gerfalk, P. J. 2006. Call for Papers: Information Systems Research Special Issue: Flexible and Distributed Information Systems Development, Information Systems Research (17:3), pp. 323325. Forte, A., Larco, V., Bruckman, A. 2009. Decentralization in Wikipedia Governance, Journal of Management Information Systems (26:1), pp. 4972. Gasson, S. 2004. Rigor In Grounded Theory Research: An Interpretive Perspective on Generating Theory From Qualitative Field Studies, in The Handbook of Information Systems Research, Whitman, M. E., and Woszczynski, A. B. (eds.), IGI Publishing, pp. 79-102. Garling, C. 2011a. Facebook Pulls Back Curtain on Timeline, Wired, http://www.wired.com/wiredenterprise/2011/12/facebook-timeline-anatomy/, retrieved March 22, 2013. Garling, C. 2011b. Facebook Auditions Kid Hackers With All-Night Codefest, Wired, http://www.wired.com/wiredenterprise/2011/12/facebook-hackathon/, retrieved March 22, 2013. Gibson, J. J. 1979. The Ecological Approach to Visual Perception, Lawrence Erlbaum Associates: Hillsdale, NJ, USA. Gopal, A., and Gosain, S. 2010. The Role of Organizational Controls and Boundary Spanning in Software Development Outsourcing: Implications for Project Performance, Information Systems Research (21:4), pp. 960982. Gray, P. H., Salvatore, P. and Iyer, B. 2011. Innovation I mpacts of Using Social Bookmarking Systems, MIS Quarterly (35:3), pp. 629-643. GroupMe. 2013. https://groupme.com/about, retrieved March 22, 2013. Guinan, P. J., Cooprider, J. G., and Faraj, C. 1998. Enabling Software Development Team Performance During Requirements Definition: A Behavioral Versus Technical Approach , Information Systems Research (9:2), pp. 101-125. Hahn, J., Moon, J. Y., Zhang, C. 2008. Emergence of New Project Teams from Open Source Software Developer Networks: Impact of Prior Collaboration Ties, Information Systems Research (19:3), pp. 369391.

17

Harris, M. L., Collins, R. W., and Hevner, A. R. 2009. Control of Flexible Software Development Under Uncertainty, Information Systems Research (20:3), pp. 400419. Hellmann, T., and Perotti, E. 2011. The Circulation of Ideas in Firms and Markets, Management Science (57:10), pp. 18131826. Huang, Y., Singh, P. V., Mukhopadhyay, T. 2012. Crowdsourcing Contests: A Dynamic Structural Model of the Impact of Incentive Structure on Solution Quality, in Proceedings of the 33th International Conference on Information Systems (ICIS), Orlando, FL, USA. Information Systems Research Calls for Papers, 2013. Special Issue on Information, Technology, and the Changing Nature of Work, https://www.informs.org/Pubs/ISR/Calls-for-Papers, retrieved May 12, 2013. Jarvenpaa, S. L., and Majchrzak, A. 2008. Knowledge Collaboration Among Professionals Protecting National Security: Role of Transactive Memories in Ego-Centered Knowledge Networks, Organization Science (19:2), pp. 260276. Jeppesen, L. B., and Lakhani, K. 2010. Marginality and Problem-Solving Effectiveness in Broadcast Search , Organization Science (21:5), pp. 10161033. Ji, Y., Mookerjee, V. S., and Sethi, S. P. 2005. Optimal Software Development: A Control Theoretic Approach , Information Systems Research (16:3), pp. 292306. Klein, H. K., and Myers, M. D. 1999. A Set of Principles for Conducting and Evaluating Interpretive Field Studies in Information Systems, MIS Quarterly (23:1), pp. 6794. Kraut, R., and Streeter, L. 1995. Coordination in Software Development, Communications of the ACM (38:3), pp. 6981. Langley, A. 1999. Strategies for Theorizing from Process Data, Academy of Management Review (24:4), pp. 691710. Lee, J., and Berente, N. 2012. Digital Innovation and the Division of Innovative Labor: Digital Controls in the Automotive Industry, Organization Science (23:5), pp. 14281447. Leonardi, P. M. 2011. When Flexible Routines Meet Flexible Technologies; Affordance, Constraint, and the Imbrication of Human and Material Agencies, MIS Quarterly (35:1) pp. 147-167. Leonardi, P. M., and Barley, S. R. 2008. Materiality and Change: Challenges to building Better Theory about Technology and Organizing, Information and Organization (18), pp. 159176. Liang, D. W., Moreland, R., and Argote, L. 1995. Group Versus Individual Training and Group Performance: The Mediating Role of Transactive Memory, Personality and Social Psychology Bulletin (21:4), pp. 384-393. Lincoln, Y. S., and Guba, E. G. 1985. Naturalistic Inquiry, Sage Publications: Beverly Hills, CA, USA. Locke, K., Golden-Biddle, K., and Feldman, M. S. 2008. Making Doubt Generative: Rethinking the Role of Doubt in the Research Process, Organization Science (19: 6), pp. 907918. Lundin, R.A., and Sderholm, A. 1995. A Theory of The Temporary Organization, Scandinavian Journal of Management (11:4), pp. 437-455. MacCormack, A., Rusnak, J., and Baldwin, C. Y. 2006. Exploring the Structure of Complex Software Designs: An Empirical Study of Open Source and Proprietary Code, Management Science (52:7), pp. 10151030. M ntyla, M. V., and Lassenius, C. 2009. What Types of Defects Are Really Discovered in Code Reviews?, IEEE Transactions on Software Engineering (35:3), pp. 430-448. Majchrzak, A., Jarvenpaa, S. L., Hillingshead, A. B. 2007. Coordinating Expertise Among Emergent Groups Responding to Disasters, Organization Science (18:1), pp. 147161. Majchrzak, A., and Markus, M. L. 2013. Technology Affordances and Constraint Theory of MIS, in Encyclopedia of Management Theory, Kessler, E. H. (ed), Sage Publications. Maruping, L. M., Venkatesh, V., and Agarwal, R. 2009. A Control Theory Perspective on Agile Methodology Use and Changing User Requirements, Information Systems Research (20:3), pp. 377399.

18

Myers, M.D. 2009. Qualitative Research in business and Management, Sage Publications: Trowbridge, UK. Myers, M. D., and Newman, M. 2007. The Qualitative Interview in IS Research: Examining the Craft, Information and Organization (17), pp. 226. NASA. 2013. The Power of Mass Collaboration, http://open.nasa.gov/blog/2013/04/22/the-power-of-masscollaboration/, retrieved April 23, 2013. Nye, J. L., and Brower, A. M. (eds.) 1996. Whats Social About Social Cognition? Research on Socially Shared Cognition in Small Groups, Sage Publications: CA, USA. Oh, W., and Jeon, S. 2007. Membership Herding and Network Stability in the Open Source Community: The Ising Perspective, Management Science (53:7), pp. 10861101. OMahony, S., and Ferraro, F. 2007. The Emergence of Governance in Open Source Community, Academy of Management Journal (50:5), pp. 10791106. Orlikowski, W. J. 2000. Using Technology and Constituting Structures: A Practice Lens for Studying Technology in Organizations, Organization Science (11: 4), pp. 404-428. Orlikowski, W. J. 2002. Knowing in Practice: Enacting a Collective Capability in Distributed Organizing, Organization Science (13:3), pp. 249-273. Orlikowski, W. J. 2007. Sociomaterial Practices: Exploring Technology at Work, Organization Studies (28:9), pp. 1435-1448. Orlikowski, W. J., and Robey, D. 1991. Information Technology and the Structuring of Organizations, Information Systems Research (2:2), pp. 143-169. Orlikowski, W. J., and Scott, S. V. 2008. Sociomateriality: Challenging the Separation of Technology, Work, and Organization, The Academy of Management Annals (2:1), pp. 433-474. Orlikowski, W. J., and Yates, J. 2002. Its About Time: Temporal Structuring in Organizations, Organization Science (13: 6), pp. 684700. Park, C., H. Son, K. H., Lee, J. H., and Bae, S.-H. 2013. Crowd vs. Crowd: Large-Scale Cooperative Design through Open Team Competition , in Proceedings of the 16th ACM Conference on Computer Supported Cooperative Work (CSCW), San Antonio, TX, USA. Pentland, B.T. 1999. Building Process Theory with Narrative: From Description to Explanation, The Academy of Management Review (24:4), pp. 711-724. Puranam, P., and Srikanth, K. 2011. Integrating Distributed Work: Comparing Task Design, Communication, and Tacit Coordination Mechanisms, Strategic Management Journal (32), pp. 849875. Sambamurthy, V., and Zmud, R.W. 2000. The Organizing Logic for an Enterprises IT Activities in the Digital Era A Prognosis of Practice and a Call for Research, Information Systems Research (11:2), pp. 105-114. Santos, A. 2013. Google Invites Glass Pre-order Holders to Hackathons in New York, San Francisco, Engadget, http://www.engadget.com/2013/01/15/google-glass-foundry-hackathon/, retrieved March 23, 2013. Sch n, D.A. 1983. The Reflective Practitioner: How Professionals Think In Action, Basic Books: New York, NY, USA. Shah, S. N. 2006. Motivation, Governance, Viability of Hybrid Forms in Open Source Software Development, Management Science (52:7), pp. 10001014. Simon, H. A. 1962. The Architecture of Complexity, Proceedings of the American Philosophical Society (106:6), pp. 467-482. Singh, P. V., and Phelps, C. 2012. Networks, Social Influence, and the Choice Among Competing Innovations: Insights from Open Source Software Licenses, Information Systems Research, Articles in Advance, pp. 122. Stewart, K. J., and Gosain, S. 2006. The Impact of Ideology on Effectiveness in Open Source Software Development Teams, MIS Quarterly (30:2), pp. 291-314.

19

Strauss, A. L., and Corbin, J. M. 1998. Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory, Sage Publications: Thousand Oaks, CA, USA and London, UK. Terwiesch, C. and Xu, Y. 2008. Innovation Contests, Open Innovation, and Multiagent Problem Solving, Management Science (54:9), pp. 15291543. Theraulaz, G., and Bonabeau, E. 1999. A Brief History of Stigmergy, Artificial Life (5:2), pp. 97116. Thomas, D. 2003. Hacker Culture, University Of Minnesota Press: Minneapolis, MN, USA. Van de Ven, A. H. 1992. Suggestions for Studying Strategy Process: A Research Note, Strategic Management Journal (13), pp. 169-191. Vidgen, R., and Wang, X. 2009. Coevolving Systems and the Organization of Agile Software Development, Information Systems Research (20:3), pp. 355376. von Hippel, E., and von Krogh, G. 2003. Open Source Software and the Private-Collective Innovation Model: Issues for Organization Science, Organization Science (14:2), pp. 209-223. Wang, P., and Ramiller, N. C. 2009. Community Learning in Information Technology Innovation, MIS Quarterly (33:4), pp. 709-734. Wegner, D. M. 1986. Transactive Memory: A Contemporary Analysis of Group Mind, in Theories of Group Behavior, B. Mullen and G. R. Goethals (eds.), Springer-Verlag: New York, NY, USA, pp. 185-208. Weick, K. E. 1979. The Social Psychology of Organizing, 2nd edition, McGraw-Hill: New York, NY, USA. Weick, K. E. 1995. Sensemaking in Organizations, Sage Publications: Thousand Oaks, CA, USA. Weick, K. E. 1998. Introductory Essay: Improvisation as a Mindset for Organizational Analysis, Organization Science (9:5), pp. 543-555. White House. 2012. Digital Government: Building a 21st Century Platform to Better Serve the American People, available at http://www.whitehouse.gov/sites/default/files/omb/egov/digital-government/digital-governmentstrategy.pdf, retrieved April 26, 2013. White House. 2011. The Obama Administrations Commitment to Open Government: A Stat us Report, available at http://www.whitehouse.gov/sites/default/files/opengov_report.pdf, retrieved April 26, 2013. Whittaker , J., Arbon, J., and Carollo, J. 2012. How Google Tests Software, Addison-Wesley: Westford, MA, USA. Yang, Y., Chen, P., Banker, R. 2009. Winner Determination of Open Innovation Contests in Online Markets, in Proceedings of the 32th International Conference on Information Systems (ICIS), Shanghai, China. Yoo, Y., Boland, R.J., and Lyytinen, K. 2006. From Organization Design to Organization Designing, Organization Science (17:2), pp. 215-229. Yoo, Y., Boland, R.J., Lyytinen, K., and Majchrzak, A. 2012. Organizing for Innovation in the Digitized World, Organization Science (23:5), pp. 13981408. Yoo, Y., Henfridsson, O., and Lyytinen, K. The New Organizing Logic of Digital Innovation: An Agenda for Information Systems Research, Information Systems Research (21: 4), pp. 724735. Yoo, Y., Lyytinen, K., and Boland, R. 2008. Distributed Innovation in Classes of Networks, in Proceedings of the 41st Hawaii International Conference on System Sciences (HICSS), Honolulu, HI, USA. Zammuto, R. F., Griffith, T. L., Majchrzak, A., Dougherty, D. J., and Faraj, S. 2007. Information Technology and the Changing Fabric of Organization , Organization Science (18:5), pp. 749762.

20