मुख्य जानकारी
Brighton-based startup GameDataCore has introduced a new platform designed to help developers, publishers, and investors gain a shared understanding of player feedback. CEO Justin French notes that industry stakeholders often make decisions based on conflicting interpretations of reality, leading to avoidable project failures.
Unlike consumer-facing tools, GameDataCore breaks reviews into classified snippets based on emotion, tone, and motivation. COO Simon Sparks explains that the system weights feedback by analyzing player profiles, including library data and playtime. This approach allows developers to distinguish between casual players and genre veterans, while also accounting for silent users who do not leave reviews.
The platform utilizes a series of small, Europe-based LLM models to ensure data sovereignty and cost consistency. Sparks emphasizes that these models avoid the unpredictability of US-based frontier models. By running specialized models locally, the company aims to provide reliable insights without relying on third-party roadmaps or external pricing structures.
Currently in testing with 24 studios since March 2026, the platform is launching alongside Steam Next Fest. Future updates aim to incorporate data from social platforms like Discord, Reddit, and TikTok, as well as internal Steamworks metrics. The tool also identifies underexploited market niches by analyzing cohorts of players who share specific, unaddressed complaints.
