Free-to-play gamers' spending habits are crucial for game developers to generate revenue in games. The unique thing about free-to-play business model games is that players don't have to pay to play the game, but need to make in-game purchases if they want a better gaming experience. The users are diverse，the trend to personalise services for users based on their diversity has become unavoidable, but identifying user needs is a large and complex problem, so an in-depth understanding of players' spending behaviour in games is essential. This thesis assists researchers in addressing this issue by constructing a game user profile based on player consumption patterns. In order to build a profile of a game's users, a great amount of data needs to be collected. However, simply providing raw data may not be of value to the development. For this reason, this thesis explores a methodology for effectively analysing and communicating survey data. This methodology collects data from user surveys, analyses the data using the author's proposed logic for creating user profiles based on consumer behaviour, and uses a combination of quantitative and qualitative methods to create a realistic game user profile. On this basis, through the application of the proposed logic, four categories of player consumption archetypes could be identified: players with Positive attitude(POSA), Moderate attitude (MODA), Reluctance attitude (RELA), and Negative attitude (NEGA). Subsequently, these archetypes could then be used to establish design guidelines for creating game mechanics and services to better encourage in-game purchases. However, as these methods have not been put into practice, the effectiveness of using them is open to question.