Professional PHP

Dave Hyatt provides a glimpse into factors influencing web page load times for safari and the mozilla based browsers. This is something that I think every web page designer and developer should be aware of.

There is a tendency in web development to focus on the amount of time it takes for the web server to generate a response for a page. Yet, this is not the largest portion of elapsed time to load a page as perceived by users. (unless you’ve messed up with server capacity). The largest time is data transfer and render time. This is especially true for users with dial-up access or older machines.

I know from personal experience that there is a strong business case for designing your pages for faster browser render times. I was able to double the average site usage per visit originating at the home page for a client’s site by optimizing the incremental load time and layout of the home page. This was an advertising supported site. More usage meant more revenue.

Because I am a rural dial-up user myself, I found the problem was that users with slow configurations or machines were simply closing the site or hitting back, rather than waiting for it to load or scrolling past the useless crud at the top of the page. Users were giving up before they reached what I’ll call the PUID: The point of useful information display.

My solution was to reduce the amount of time it took to reach the PUID. I structured the page for incremental rendering times so that browsers could render the portion “above the fold” as quickly as possible and made sure that space held useful information. The content and total load time for the page did not change much, only the layout and underlying HTML, Javascript and CSS.

I am a server application developer, not a web designer, so perhaps this type of stuff is old hat for the really good designers. Unfortunately, I have not found an abundance of information such as Hyatt is offering and certainly no comprehensive treatment of the topic.