With the news that the pageview as a relevant web metric may become another statistic (sorry) and be replaced by time on site, I was randomly inspired to look at the bmw.co.za webstats from a slightly different angle - time on site versus connection speed.

The interesting thing (for me, that is), is that the average time on site is fairly similar across all connection speeds. Does this mean that people with faster connection speeds are viewing more pages than people on dial-up? Again, the stats are quite similar with DSL only beating Dial-up by just under 1 page per visit (average).

I put the similarity of the stats down to the fact that the site loads nice and quickly, users come in and get what they need, and get out.

I thought that the M3 campaign site would show different results as it’s a lot more intensive on the bandwidth side of things. But alas, no real change in time v speed challenge. Average pages/visit are harder to rely on as the site is mainly flash based (therefore just the one “container” page loads).

Am I missing something here, or is line speed just not an influencer in a site of this nature’s usage?