Abstract
Search Engine Optimization (SEO) is important for improving the visibility of websites in search results, but checking every page of a website manually is slow and difficult for non-technical owners. To address this problem, this paper proposes an ML based SEO performance analysis and recommendation system built using Python and Flask. The user enters a website URL, and the system crawls the accessible pages using Requests and BeautifulSoup and extracts six important SEO features: content length, keyword density, internal links, external links, meta description and image alt text. An SEO score out of 100 is calculated using predefined rules, and the strengths and weaknesses of the website are identified. A Random Forest classifier trained on an SEO dataset predicts whether the ranking-improvement outcome is likely to be positive. Based on the analysis, the system recommends actions such as adding relevant content, improving internal linking and including missing meta descriptions and image alt text. The results are stored in a MySQL database and displayed through a user-friendly web interface, helping website owners to understand common SEO issues and identify possible improvements without checking every page manually.