Управління архітектурою SDK для розробки мобільних застосунків у контексті конфіденційності маркетингових даних

Mobile applications increasingly rely on embedded analytics software development kits (SDKs) to monitor usage, optimize user experience and support data-driven business decisions. At the same time, empirical studies show that many third-party SDKs exfiltrate privacy-sensitive data, misrepresent their practices in privacy policies and expose applications to supply-chain security risks. These findings create a tension between the need for fine-grained behavioral data and strict requirements of regulations such as the GDPR and CCPA. The paper proposes a privacy-aware reference architecture for a mobile analytics SDK intended to be integrated into native applications as a first-party or carefully governed third-party component. The architecture decomposes the SDK into five layers: Core Module, Data Collection & Processing, Storage & Queuing, Networking & Communication, and Integration. For each layer responsibilities, data flows, and embedded privacy controls, including consent-driven initialization, configurable event schemas, local pre-processing, encrypted transport, throttling, and auditable public APIs were specified. The research design combines analysis of recent empirical studies on SDK privacy risks and platform policies, conceptual modeling, and inspection of configuration options in contemporary analytics SDKs . The proposed model is then discussed against existing evidence about default SDK behavior, fingerprinting practices, and developer constraints. The contribution of the paper is twofold: it systematizes scattered knowledge about how mobile analytics SDKs operate, and it offers a practical blueprint that helps developers align client-side analytics with modern privacy and security expectations while preserving the analytical value of event data.