"Years of building, pitching and raising for games made the same failure painful to watch" – Why GameDataCore wants to find out how players feel
GameDataCore launches a platform using Steam reviews to help developers interpret player feedback and prioritize fixes, timed for Steam Next Fest.
GameDataCore, a Brighton-based start-up, has launched a platform designed to analyze Steam reviews and extract meaningful insights for game developers. The tool aims to address the challenge of interpreting player feedback, which can be inconsistent or misleading when condensed into simple metrics like thumbs-up or thumbs-down ratings. By breaking down reviews into detailed snippets, the platform classifies feedback across topics, emotions, and motivations, providing developers with a clearer understanding of what players truly feel.
The platform’s approach involves weighting reviews based on the profile of the reviewer, such as their gaming library, playtime, and preferences. This method helps distinguish between complaints from inexperienced players versus seasoned veterans, ensuring that feedback is interpreted in context. Additionally, the tool identifies silent players—those who played but never left a review—by analyzing their engagement patterns, offering a more comprehensive view of player sentiment.
GameDataCore’s platform is launching to coincide with Steam Next Fest, an event where hundreds of game demos will be available. Developers can use this period to assess whether their intended player experiences align with actual player reactions. The platform provides a prioritized list of potential issues to address, complete with evidence from reviews and player profiles, enabling teams to share findings via Slack or Discord for collaborative decision-making.
The platform relies on a series of small, European-based LLM models to process reviews, ensuring data sovereignty and cost efficiency. GameDataCore avoids using US-based frontier models due to concerns about data privacy, consistency, and dependency on external providers. A pilot version has been tested with 24 studios and publishers since March 2026, and future updates will incorporate additional feedback sources like Discord, Reddit, and Twitch.