Paths Leading to Success

One question I hear from time to time from Omniture SiteCatalyst users is: “What are the paths visitors take on my site when they are successful in reaching one of my desired Success Events?” Basically, on most websites, there are a few ideal paths that you wish all visitors would take. Unfortunately, most visitors don’t adhere to your ideal path so you need to see what paths they are choosing to take. Of course, you can use Pathing reports to see all of the paths that visitors are taking, but that can be very cluttered and hide what is really happening related to Success Events. Therefore, one interesting analysis you can do is to focus only on paths that do lead to success. I equate this to the old “cow path” story where architects would wait to create sidewalks in front of new buildings to see where people naturally walked and then build the sidewalks where people are already going. In this post I will show you how you can do this in SiteCatalyst.

But I Don’t Have Discover or Insight?
When I raise this topic, the first thing people say is that they can’t do this because they don’t have Omniture Discover or Omniture Insight. While those products are great and I highly recommend them, the good news here is that you don’t need those tools to see the paths that lead to success. In fact, due to a current Discover limitation (more info on this later on), you cannot do this analysis in Discover, but you can do it easily in Omniture Insight. The one thing you will need to perform this analysis, however, is Advanced Segment Insight (ASI).

So How Do I Do It?
The key to this solution is the Full Paths report in SiteCatalyst. Normally, this report is a tad boring since there are so many unique paths on a site that beyond ENTER>>HOME PAGE>>EXIT, the rest of the percentages are usually pretty slim (i.e. .1% of all paths). However, if you could look at this report for a small subset of your website audience, the percentages should go up dramatically. This can be done with ASI, where you create an entire new data set for a specific segment. Once you have created your segment and run your ASI slot, you have access to all SiteCatalyst reports, including the Full Paths report. Hence, all you need to do is to properly identify a segment containing a Success Event, run your ASI slot and then look at the Full Paths report.

The way I start this is by identifying the Success Event that I care about. Let’s say that you want to see a segment of “Purchasers.” In this case the only Visits you want to include in the segment are those that have completed an Order within the Visit (feel free to review Segmentation in my old post):

Once you have created this segment, you now create an ASI slot for, say, 30 days of data (amount of days is up to you). When the ASI slot has completed processing, you can open the Full Paths report and see which paths are the most popular and include an Order:

I have found it fascinating to extract these paths and look for common trends. It is often like finding needles in a haystack! The best part, is that after you have learned what you want to learn, you can delete the ASI slot and re-create it for a different segment. Here are a few examples you can create:

• View all paths that led to an Order, but only when the visitor was a first time visitor
• View all paths that led to an Order, but only when the visitor was a return visitor
• View all paths that led to an Order that was greater than $100
• View all paths that used a Product Finder and led to an Order
• Etc…

Just remember the formula. Create segment, run ASI, look at Full Paths (Repeat). The possibilities are only limited by your imagination. I usually look at the top five paths for a bunch of different segments and then compare them to see what is common between them.

As I mentioned earlier, it would be much easier to do this analysis in Discover (where segmentation is instantaneous), but currently, there is no Full Paths report in Discover (see my idea suggestion and feel free to vote for it!). I hope that gets into the product since then you wouldn’t have to wait a few days each time for the ASI slots to process.

What About Failure?
But wait…there’s more! As Avinash Kaushik likes to say, one of the cool things about the web is that you can fail faster (and learn)! Therefore, you shouldn’t just look at success, but you should also look at failure. Why not create an ASI slot for visits where people did NOT succeed (create an Order in this case)? Find out where people are most often taking wrong turns and maybe you can correct them. Remember, when you perform normal Pathing analysis, you are seeing the good paths mixed with the bad paths, but with this solution, you can easily segment them out and analyze each separately. Here is a sample segment definition you can use (notice “Exclude” – opposite of the first segment):

Well…there you have it, a quick and dirty way to see which paths lead to success (or failure). Enjoy!