This tool uses a client-side library (pdf-lib) to optimize PDF structure by enabling object streams and stripping common metadata. These methods can reduce file size. The actual compression achieved will vary significantly based on the original PDF's content, how it was created, and the amount of metadata it contained. For PDFs with large, unoptimized images or complex structures that weren't using object streams, the reduction might be more noticeable. If a PDF is already well-optimized, contains minimal metadata, or its size is primarily due to highly compressed images (e.g., JPEGs in a scanned document), the reduction in file size may be minimal or even 0%. This is an expected outcome for such files and not an error in the compression process. This method provides good general structural optimization but does not perform deep image re-compression (like reducing image quality/DPI) or other advanced lossy optimizations often found in specialized server-based tools.