{{Multiple issues|one source = November 2011|peacock = November 2011|refimprove = February 2011}} {{notability|Web|date=August 2011}}
Web performance optimization, WPO, or website optimization involves ongoing monitoring and testing of websites to achieve optimum performance under given constraints. Usually optimization is restricted due to lack of complete information and metrics to evaluate the performance of a website.
In recent years,[when?] and starting since about 2004,[citation needed] web developers have been spending increasingly more time optimizing the speed at which websites load. The goal of web performance optimization is to drive greater traffic to a website. Until recently[when?] it was believed that page load time was determined by the server; however research at Yahoo and Google prove otherwise.[citation needed] It is now accepted[by whom?] that 80-90% of load time issues occur at the front end and web performance optimization is critical.[citation needed]
Web performance optimization tools allow monitoring of internal applications, first mile performance,[clarification needed] and measuring of quality and reliability of streaming media the way users view it. One can also access tools that provide front-end visualization of web performance real-time through a central console, conduct root cause analysis and adopt diagnostics solutions for websites and applications. These web performance optimization tools allow for rapid correction of errors so that users get a seamless, high quality web experience.
There are numerous factors influencing the performance of a web page over the Internet. Factors that are explicitly studied to optimize web performance are browser rendering efficiency, protocol options, response-request cycles, caching, multi-server content and byte range requests.[citation needed] Some of the factors that are usually not included in the studies are DNS lookup time, redirection, dynamic content and packet level performance.[citation needed]
The main factors accounted for in testing and analysis is network delay, server load and number of objects and total byte.[1]