If your data can be grouped into a discrete sequence of events
(step 1: autotrader.com, step 2: used cars, step 3: acura tl ...)
then you could maybe get a high level visualization of all users
using a Jump Plot:
Many great things about this visualization:
show all different userpaths taken in one view,
can plot many separate lines-one for each user,
Can you show me how it is to be done as I am unaware.
Sorry I didn't look closely at the data.
The jump plot is more if there is a set sequence of actions.
I don't think it will work here. In this case it looks like users
can go to a wide range of possible urls.
To take a step back, what are the questions you are seeking to answer?
how many times was each url visited, and show it by size?
how many urls did an individual peruse?
what urls are most often viewed together?
Attached in the forum post is an example showing
all the urls as a bubble chart, with size encoded as
the number of views, model-make encoded as color.
Upon clicking on the url bubble, it filters the lefthand
graph to just those profile IDs that viewed the page.
The lefthand graph shows each ProfileID as a row
and the sequential Page Paths that they viewed.