Research

Platform operators face a continuous trade-off in governance decisions between addressing user interests and preserving incentives for complementor innovation. One highly debated area of platform governance concerns protecting user privacy by restricting third-party tracking, which, however, is vital to many complementors’ value capture models. In this paper, we seek to understand the implications of stricter privacy governance on complementor innovation. We examine how mobile app developers adjust their innovation activities in response to Apple’s App Tracking Transparency (ATT) policy. Using a quasi-experimental difference-in-differences framework, we compare the innovation outcomes (feature updates) for apps affected by the policy with those unaffected, before and after the policy announcement. Our results indicate a positive effect of ATT on affected apps’ innovation relative to unaffected apps after the policy announcement. Further analyses are consistent with an adaptive response to reduced third-party tracking: the innovation response is stronger among more exposed apps, and affected developers innovate to support alternative ways to capture value. We discuss that while stricter privacy governance may initially seem to negatively affect complementor innovation, its nuanced effects can also serve as strategic catalysts and enable platform operators to reshape complementors’ innovation activities.

Measuring innovation often relies on context-specific proxies and expert evaluation. Hence, empirical innovation research is often limited to settings where such data is available. We investigate how large language models (LLMs) can be leveraged to overcome the constraints of manual expert evaluations and assist researchers in measuring innovation. We first design an LLM framework that reliably approximates domain experts’ assessment of innovation from unstructured text data. We then evaluate the performance and broad applicability of this framework through two demonstration studies in different contexts: (1) innovation in software application updates and (2) user-generated ideas in product reviews. We compared the performance (F1-score) and reliability (consistency rate) of our LLM framework against alternative measures used in prior innovation studies, and to state-of-the-art machine learning- and deep learning-based models. The LLM framework achieved higher F1-scores than the other approaches, and its results are highly consistent (i.e., results do not change across runs). This article equips R&D personnel in firms, as well as researchers, reviewers, and editors, with the knowledge and tools to effectively use LLMs for measuring innovation and evaluating the performance of LLM-based innovation measures. In doing so, we discuss the impact of important design decisions—including model selection, prompt engineering, training data size, training data distribution, and parameter settings—on performance and reliability. Given the challenges inherent in using human expert evaluation and existing text-based measures, our framework has important implications for harnessing LLMs as reliable, increasingly accessible, and broadly applicable research tools for measuring innovation.

The mobile apps that mediate everyday activities make data disclosures a routine part of people’s lives, but in fragmented prior research, the privacy threats associated with data disclosures dominate. To complement such perspectives, the current research proposes a novel conceptualization that reflects how consumers respond when transparency policies make data disclosure practices salient. A large-scale, natural experiment involving more than 700,000 app reviews, surrounding the introduction of Apple’s Privacy Nutrition Labels, reveals a negative relative effect of transparency on consumer attention to data disclosures. Yet as these labels diffused, and transparency became a platform standard, consumers seemed to attend specifically to data disclosures by apps that were very late in adopting the labels. In contrast with prior evidence, the overall sentiment in data disclosure reviews remains largely positive. To clarify these findings, two controlled experiments systematically vary the timing of transparency efforts and the appropriateness of the data disclosures. The experimental evidence reveals that consumer reactions depend jointly on whether transparency occurs when expected and if the disclosed data practices align with contextual expectations. Thus, consumer responses to data disclosures are shaped by their timing and appropriateness, not just the simple existence of data disclosures.