Select data structures based on expected collection size and usage patterns. For small collections (typically fewer than 30 elements), consider using ordered containers like `std::map` instead of hash-based ones like `std::unordered_map`. With small collections, the theoretical O(log n) vs O(1) lookup time difference becomes negligible in practice, while ordered containers avoid hash function computation overhead and potential collision handling. ```cpp // Instead of using unordered_map for small collections: std::unordered_map module_requests; // Consider using an ordered map: std::map module_requests; ``` When working with hash-based containers, reuse precomputed hash values when available (like V8's `GetIdentityHash()`) instead of recomputing hashes with `std::hash`. This avoids unnecessary computation and improves performance, especially for string data that might already have an identity hash from the JavaScript engine.