Friday, December 25, 2009

Path Analysis : Does it makes sense ?

Basic idea about Websites : There are certain success paths for the website. Whereas this may be true for a real market scenario, this is not relevant for websites, at least for most of them, the reason being most of them are unstructured i.e. there are a vast number of paths or ways in which a user can navigate to the conversion or goal of the website. It becomes almost impossible to track almost all of them , particularly with the way in which users navigate through them.

So why do website creators want to create a particular path that they would like the users to follow?

* They want to take the users through to the final goal of the website.(shopping cart conversion, lead conversion)
* They want to control the user experience on the website

A typical path analysis may not end up to be a sound ROI for most websites.


Again for most websites we have a few landing pages as we go ahead in to the flow this bifurcates into a large number of paths, because of the choice of navigation available.

In most cases a typical path analysis report shows that users don't end up following the usual/expected course set out.

Only 2% of the users tend to follow the predefined path flows that the website owner expects. The rest follow a jumble of path flows that cannot be traced linearly across the website.

In all such cases path analysis seems like a sheer waste of time and money .

This behavior in website traffic can be primarily attributed to the traffic referrals from the search engines. In case of most websites 20-30 % to 80-90% of the incoming traffic is determined by the search engines and as such the home page/landing page is actually not determined by the website owner, but by the search engines. So there is no such thing as a optimal path , and especially in a world dominated with search engines it's a lot more sub optimal .


If you have a website which sells say a book or a car etc, and suppose there are 5 pages 1,2,3,4,5 where you expect the user to make the purchase on the 5th page. Rarely would any user follows this linear path. In reality the path can be 1,2,3,2,3,1,4,5.

However the current set of analysis tools cannot accommodate for this back and forth behavior. They expect people to go down a particular linear path(1,2,3,4 and finally 5).

Browsing behavior in the web is generally not linear, however it is only linear behavior that the current path analysis tool can track

Rich Applications:

Web Applications have evolved from Static Web pages to Dynamic content to Rich Internet applications(using AJAX, Flex, Flash).

Static--> Dynamic-->Rich….

You are creating a rich experience for each user. So it becomes worthless to do a path analysis.
Exception to the above rule are the websites where you get Structured experiences.

Structured experiences:

Examples including website where there is only one possible flow path. Path analysis in all such cases can be considered to be more of Abandonment Analysis. Consider the case of a Credit Origination Application where a user moves through an application form filling up specific details and finally end up on the "Thank You" page .

Alternatives to Path Analysis: Funnel Reports:

Funnel report break down the website pages in to Logical Content Groups of Influence, which helps to identify the movement of traffic from one group to the next. They help to concentrate on the macro level usage of the website and break down the website in to well defined logical funnels. It also helps to identify the pages in each group that contribute the most to the site's conversion and necessary changes to all such pages can be considered as actionable insights and as such, these analyses are much better ROI for the investor's money.

No comments:

Post a Comment