For a frm in an open source software (OSS) ecosystem, the requirements engineering (RE) process is rather multifaceted. Apart from its typical RE process, there is a competing process, external to the frm and inherent to the frm’s ecosystem. When trying to impose an agenda in competition with other frms, and aiming to align internal product planning with the ecosystem’s RE process, frms need to consider who and how infuential the other stakeholders are, and what their agendas are. The aim of the presented research is to help frms identify and analyze stakeholders in OSS ecosystems, in terms of their infuence and interactions, to create awareness of their agendas, their collaborators, and how they invest their resources. To arrive at a solution artifact, we applied a design science research approach where we base artifact design on the literature and earlier work. A stakeholder infuence analysis (SIA) method is proposed and demonstrated in terms of applicability and utility through a case study on the Apache Hadoop OSS ecosystem. SIA uses social network constructs to measure the stakeholders’ infuence and interactions and considers the special characteristics of OSS RE to help frms structure their stakeholder analysis processes in relation to an OSS ecosystem. SIA adds a strategic aspect to the stakeholder analysis process by addressing the concepts of infuence and interactions, which are important to consider while acting in collaborative and meritocratic RE cultures of OSS ecosystems.
Firms that use open source software (OSS), e.g., as part of their supporting infrastructure, product strategy, or business model, need to consider the requirements engineering process of the OSS itself . This second, external to the focal frm, RE process is facilitated by the software ecosystem (cf. OSS community ) that surrounds the OSS . Firms that are users of the OSS may also be involved in its development and maintenance and can be considered as members of the ecosystem, as well as stakeholders to the OSS. We refer to Glinz & Wieringa’s defnition of a stakeholder as “...a person or organization who infuences a system’s requirements or who is impacted by that system” . In our context, we consider a person or an organization as the members of an OSS ecosystem, and the system being the OSS that underpins the ecosystem, using the defnition by Jansen et al .
RE practices in OSS ecosystem may be described as informal and decentralized. There is often no central repository with requirements defned in the problem space, describing the product of need, along with heavy processes and tools for examining the requirements for completeness and consistency . Instead, RE may be considered as a lightweight and evolutionary process of requirements refnement . Practices such as elicitation, specifcation, and prioritization overlap and are done collaboratively through iterative and transparent discussions and negotiations including up-front implementations [6–8]. These discussions and implementations of requirements are spread out over a number of requirements artifacts, each with its own repository. Examples of these artifacts (cf. informalisms ) include reports in an issue tracker, messages in a mailing list, or commits in a version control system. Prioritization is commonly conducted by stakeholders with central positions in the ecosystem’s governance structure [9, 10]. To gain such a position in OSS ecosystems with a meritocratic governance structure, a stakeholder needs to prove merit by being active, contributing back, and having a symbiotic relationship with the OSS ecosystem .
Hence, the focal frm is one stakeholder among many within an open and fuctuating population in the OSS ecosystem . This can result in conflicting agendas and lack of control, e.g., in regard to which requirements to be implemented and prioritized, render misalignment with internal RE processes , and complicate contribution strategies . The focal frm may, therefore, have to gain the infuence necessary to afect the RE process in an OSS ecosystem according to its own agenda.
The Merriam-Webster dictionary1 defnes infuence as “the power to change or afect someone or something”. In our context, this relates to the power of a stakeholder to change or afect the RE process in an OSS ecosystem. This notion of infuence aligns naturally with what defnes a stakeholder , and as a characteristic enables frms to, e.g., see the requirements in which stakeholders hold a certain interest, and from there be able to create an overview of their agendas in the ecosystem . Further, this understanding enables the focal frm to analyze how these stakeholders invest their resources in order to satisfy their agendas . By also considering other stakeholders’ interactions within the ecosystem, frms may identify possible partners and competitors . Moreover, this can help frms to learn how to adapt their own strategies and processes with the OSS ecosystem’s and how to build their own infuence and position the ecosystem’s governance structure . The knowledge output can then be leveraged toward other stakeholders through the politics and negotiations that take place in the ecosystem’s RE process .
These aspects highlight the importance of stakeholder identifcation and analysis as input to the continuous and complex decision-making process which RE constitutes  by helping to answer questions as which other stakeholders exist in the ecosystem, what are their agendas, and how do they aim to achieve them . However, current practices  are not adapted to consider these strategic aspects  in the context of OSS ecosystem  and its informal and collaborative RE process [6, 7], specifcally the importance of understanding stakeholders’ infuence and interactions. Involved frms are no longer the vantage point, and instead, form part of a larger set of interdependent stakeholders . We address this gap with a design science research approach [20, 21] and defne it as a design problem :
DP How to characterize the infuence of stakeholders on the OSS ecosystem’s RE process, so that a focal frm can understand other stakeholders’ agendas, collaborations, and resource investments in pursuing these agendas?
The contribution of our work is the proposal of the stakeholder infuence analysis (SIA) method. Its aim is to help frms to analyze an OSS ecosystem to identify its stakeholders’ infuence by the impact they have with respect to the requirements that get implemented in the OSS. We base SIA on social network analysis constructs [22–24] that have proven to be useful in characterizing the infuence of stakeholders [15, 25], but also efective when analyzing a frm’s participation in OSS ecosystems [25, 26] and requirement-centric stakeholder collaborations [27–29]. An analysis approach used in an earlier reported case study of the Apache Hadoop OSS ecosystem  is formalized to consider how requirements may be informally represented in multiple artifacts in decentralized repositories present in OSS ecosystems [6, 7]. The infuence analysis is then operationalized with a stakeholder mapping approach based on earlier work [31–33]. To demonstrate SIA’s applicability and utility, we present a case study of the Apache Hadoop OSS ecosystem.
The rest of this paper is structured as follows: In Sect. 2, we describe the research approach used in the development of SIA. In Sect. 3, we give a detailed presentation of SIA, while in Sect. 4, we demonstrate its applicability and utility with a case study. In Sect. 5, we discuss alternative approaches to characterize infuence and threats to validity. Finally, we conclude the paper in Sect. 6.