Etsy has made significant strides in improving web performance by utilizing the Speculation Rules API (SRA) on its search pages. This collaboration between the search and web performance teams led to 20-24% enhancements in loading times for product listing pages, with some metrics even dropping to 0 milliseconds. The SRA allows for more efficient resource fetching through methods like prefetching and prerendering. Prefetching downloads resources that are likely needed soon, while prerendering loads and displays pages in the background, offering faster navigation when a user clicks. One challenge Etsy faced is that its HTML pages are non-cacheable due to their complexity. To maximize the benefits of prefetching, they implemented a strategy to keep prefetched pages in memory for five minutes. This is crucial because once that time expires, standard caching rules apply. Enhancements were also made to manage link attributes, as JavaScript modifications on hover created issues with prefetching. If the href of a link changes during hover, the browser may abandon the prefetching process, which can lead to inefficiencies. Analytics also posed a challenge during this implementation. Keeping the tracking data accurate after introducing the SRA required careful adjustments. The team spent considerable time ensuring that their analytics systems remained reliable while accommodating the new prefetching strategy. Overall, the effort illustrates a methodical approach to optimizing performance while navigating the complexities of modern web development.