eValid scripts were used to establish actual "below the fold" timing data for a popular web search engine. After analysis of many weeks of data the customer made changes in their site structure that significantly improved response times and result customer success rates.
Most web search engines normally have a part of the page that is displayed before the search results are displayed. The fixed part is called "above the fold" and the variable part where the results are shown is called "below the fold". Two examples are Google or Ask, but most search engines have this structure.
The end-user's response time "below the fold" is an indicator of both the speed of the web connection and the performance of the search engine computer complex. Having a quick response time
eValid Application Description
To measure the response time "below the fold" eValid was programmed (through the script language) to reset its internal stopwatch timer at the instant the user clicks "Search" -- after the target of the search has been entered. Then, eValid enters into a wait loop that determines how much time is required for (1) the first results to show up and (2) the entire first page of results to show up. This is done by synchronizing [and timing] on pre-set text items in the response display.
eValid Capabilities Exploited
This project used eValid's internal timing capability (the built-in event stopwatch) as well as advanced modes of page synchronization including synchronization of text and visual page components.
Regular eValid measurements of response time showed consistent patterns and variations of the end-user "below the fold" response times.
After analyzing eValid data our customer made a number of changes to the support structure for their search engine including changes to the load balancers, modifications to the geographic locations and coverage, and certain internal improvements.
The result of the changes was faster, more-quickly displayed search results pages that were less likely to encourage "user click away" because of response time problems.