This research project is about sustained engagement of third party app developers in platform ecosystems. Platform ecosystems consist of a core product, interfaces that allow for extensions, and complementary products and services (e.g., apps) that are interoperable with the product (Baldwin & Woodard, 2009; Tiwana, 2013). The creation of complementary products is often the result of independent third party development (e.g., Boudreau & Jeppesen, 2015). As such, the value for users is much dependent on such complements as they expand the functionality of the platform (e.g., Boudreau, 2012). As a platform owner it is therefore key to attract a large pool of skilled app developers. However, besides bringing developers ‘on board’, we miss a framework on how to keep developers engaged over time.
Keeping developers involved over time is important because when developers withdraw their involvement with the platform, their apps become obsolete and hamper end-users in using the platform. App developers leaving the platform is a recognized problem that deserves more attention (Tiwana, 2016). Since we lack research on third party app developers specifically, we draw on related research in the user innovation literature, and open source software development in particular, which has extensively theorized why individuals develop and contribute (e.g. Von Krogh et al., 2012). Yet why developers decide to continue, change, or discontinue development is less well studied. In addition, in open source software projects, another developer can take over where the previous has left off, which makes the withdrawal of a developer in the open source community less problematic. It is, however, questionable whether these findings apply to third party app developers: each unique app is developed by a particular developer which makes the performance of that app the sole responsibility of that particular developer.
We study the Philips Hue smart lighting system, which consists of consumer LED light bulbs that can be controlled through an app on a smartphone or tablet. Besides the official Philips app, over 400 third-party applications have been developed on the Hue platform. Given the lack of well-defined concepts and theories on the sustained engagement of third-party app developers, we followed an exploratory field study (Edmondson & McManus, 2007) approach and gather in-depth qualitative data. The first author is an embedded researcher at Philips, where she observes meetings and has conversations with the Philips Hue team. To investigate the engagement of various developers over time, we held interviews with third party app developers.
From our preliminary analyses we find that an app development process involves different tasks with different associated motivations. We distinguish broadly between creating and maintaining, where creating involves actions regarding the initial development of the app (i.e. writing the code, exploring the API and documentation) and adding functionality to the app over time (additional features), and maintaining the tasks of fixing bugs, making the app work with new operating system versions as well as providing user feedback and promoting the app.
Developer engagement then involves a developer’s commitment to both creating and maintaining over time. Our analysis shows that developer engagement can follow different trajectories: 1) a developer creates an app but over time leaves the platform, 2) a developer creates an app and carries out the minimal required effort to maintain the app, and 3) a developer creates an app and stays actively involved with the platform through continuous maintenance and creation. In the latter case, developers can opt for continuous creation within one app (adding features over time) or through developing multiple apps. Furthermore, a developer may decide to have the same app on multiple platforms (e.g. both iOS and Android), and thereby deliver the same functionality but still ensure differentiation as they address different user bases.
We add to the literature on developer motivations by studying the specific setting of app development. Our results confirm that developers are driven by a particular user need (e.g. missing functionality), financial gains or an interest in new technology and the enjoyment from ‘tinkering’. In contrast with findings from other research settings, we did not find developers indicating that signaling, reputation or reciprocity towards other developers is important. However, we discovered that end-user feedback is an important mechanism that drives developer engagement. That is, end-users requesting and suggesting new features that the developers then integrates, as well as a sense of responsibility so that end-users are not left with a poorly functioning app.
Our findings have important implications for how platforms owner actions can affect developers’ decisions to stay engaged or leave. Platform owners can help developers in taking over tasks that they struggle with and as a result fail to do otherwise. For instance, the Hue case shows that developers struggle with marketing their app, which is important to attract end-users. In our case, Philips has created an “More apps for Hue” list. This ‘app gallery’ makes the apps easier to find for the end-users, and result in increased downloads, which is an important driver for platform engagement. Furthermore, platform owners can introduce new hardware or software to keep developers interested over time as this provides new opportunities for tinkering and extending the app functionality. At the same time, platform owners can (unintentionally) discourage developers to further pursue app development, for example, when the platform takes over functionality and integrates these in the platform.
Lastly, we unpack what a valuable contribution is in app development, and show that only through sustained developer engagement complementary value is created in platform ecosystems. At minimum an app developer needs to stay engaged to do the required maintenance to his app(s). At the same time, abandoning an app does not necessarily have to problematic, for example when an app becomes obsolete. More importantly, if the same developer stays in engaged through another app, he is more likely to provide a better service to the end-users and platform as a whole as he benefits from learning effects. In conclusion, our paper provides insight in developer engagement by showing how and why third party app developers decide to stay or leave a platform.
