The first person I heard criticize the conversion funnel model was Jeff Eisenberg in July/August of 2005. He told me then that while analytics vendors like WebSideStory, Omniture and CoreMetrics all measured funnels (the step conversions between a product page and a purchase point) that they weren’t adequate because they were very linear.
By linear he meant that the tools only measured a single
path (in a typical conversion funnel – shown right).
This meant if the customer left this path at any point on the process, the systsems didn’t adequately show where and how, unless you were setting up multiple conversion funnels to cater for every possible interaction, which sounds good but in reality isn’t possible.
I largely agreed, but the case is that most tools can’t measure anything better than a defined funnel at a time. Unless you have plenty of booty in your treasure chest and you can afford something like Visual Sciences (now part of WebSideStory) which gives you much clearer insight as to where people leave and why.
This article however from Shane Atchison (CEO of ZAAZ), at least describes a methodology around which new scenarios and funnels could be built. Shane describes a system he is calling “The Hub And Spoke Model” where a product page is placed in the hub and the links around it could be tracked. Again something like Visual Sciences already measures this activity in an effective way (if I understand this concept right), but for products like Google Analytics, CoreMetrics, Omniture, and HBX (without VS) this could be an excellent method to use to deliver more insight as to where and what people are doing after visiting your site.