As I read through Understanding How Users View Application Preformance one thing was very clear to me. Not all pages are created equal. The article author was attempting to determine how long a page should take to load for a user to be satisfied, and tried to reconcile all the different information on this, from IBM’s one second rule for application response to forester’s 30 second rule for page load to Spool’s studies on perceived load time.
What if they are all right? A page that is an article is not the same as a page that is part of a application process in the eye of the user, even if they are made of the same raw materials–html. When you are in the middle of a process… perhaps setting up an apointment in an online calendar or editing an address in an online mail program… waiting between steps is excrutiating.
But when you are shopping or reading an article, asking for a page of data and waiting for it is far less painful.
In my experience a user would rather wait for a page that holds all they need to know– fabric swatches, size choices, etc– than have to click through many small pages to get the same infomartion.
That changes again upon checkout, when you are hoping to speed through the task and the same 20 second wait to see all the chino choices you didn’t mind before suddenly seems an eternity as you go from shipping address to billing address.
This obviously has repercussions for design. Multipage tasks should be composed of sleek fast loading pages; information pages are worth waiting and should hold all the information a user seeks.