How many times have you heard the term "User Path Analysis" being mentioned as something you should conduct for your site? If you have anything to do with managing a Web site or Web analytics, chances are, you have considered it... or maybe gone down that path!
Pathing is, as expected, the path a visitor takes through a site. It is literally a flowchart of how a visitor makes his way from the landing page to the exit page and the content consumption between these two.
It does, thus, seem intuitive to want to analyze this user path -- to determine the most common path users take before a desired outcome (a purchase, a request for information, etc.). This information can then be used to modify site navigation or copy to push visitors down that trusted, successful path. This approach, however, is not as efficient as it sounds.
For one, consider the number of paths a visitor could take. An example:
Path A: "Home page" to "Browse store" to "Clearance" to "Kids" to "Home page" to "Furniture" to "Kids" to exit
Path B: "Landing page (from search engine)" to "Kids" to "Clearance" to "Home Page" to "Furniture" to "Kids" to exit
Now consider a 100- or 200-page site and the possible number of paths a visitor can take. Keep in mind the complexity that the back button and page search add. There are simply too many options. Web consumption is not linear and neither are user paths. This means that there probably isn't a magic path to conversion.
Also, due to the wide variety of options, there is a small percentage of visitors that actually follow a common path. If only 5% of visitors follow the same path to a purchase, would it be wise to try and push the other 95% down that path as well? And even if we want to, is it doable? Probably not.
This isn't to say that there is no merit in looking at this information. However, it needs to be managed carefully. We need to address the issues of linearity and too much information.
Hence, user path analysis has direct applicability in the linear sections of the site -- e.g., checkout pages. If all the consumer is expected to do is hit "next," user path brings to light the drop-off points in the purchase funnel -- for example, most prospects drop off between the "Choose your color" and "Review order details" pages. Why? Is the color selection not right? Not adequate? Is the next page not loading quickly? Etc. etc.
The information can also be used to answer a specific question. In the first example - Are users interested in buying kids' products also looking for furniture deals? Such specific questions make the data and the analysis manageable, meaningful and actionable.
Another way of gaining these insights is through segmentation. For example -- all visitors browsing the kids' pages are one segment. This means that we don't need to worry about the specific path they take within the kids' section, how many times they hit the back button, etc. We start to analyze the data when they leave the kids' section -- what other products are they interested in, what site features they navigate through and what products they come back to buy. Of course, this segmentation varies by purpose of the site and its functionality.
As attractive as it might sound, don't jump right in when someone wants to look at user path. Take the time, understand the question being asked and make your dataset manageable. Analysis paralysis isn't uncommon on this one.