If we follow a star schema, will it be better than an optimized or snowflake schema? But that way it will take more space, right? For example, the same data will be repeated multiple times if it is not normalized.

A star schema is a denormalized design where a central fact table is surrounded by dimension tables. Because the dimensions are flattened, queries can read the fact and dimensions in a single join, which is fast and easy for BI tools. The trade‑off is that dimension data is duplicated across many rows, so the schema consumes more storage. A snowflake schema normalizes the dimensions into sub‑tables, reducing redundancy and saving space, but each query must perform additional joins to reconstruct the full dimension. In practice, a star schema is often preferred for reporting because query performance outweighs the extra storage cost, especially when the data volume is moderate and the BI workload is read‑heavy. If storage is a critical constraint or the dimensions are very large and rarely queried, a snowflake or hybrid approach may be more appropriate. Thus, whether a star schema is “better” depends on the specific workload, storage budget, and performance requirements.