Friday, March 8, 2013

The power of session-level analytics reporting in Sitecore Analytics

Some background on session-level analytics in Sitecore:


The conversation often comes up when discussing Sitecore Analytics: "when should I use Sitecore Analytics and when should I use xyz Analytics?".  The use of multiple analytics tools, the coordination between them, and/or the decision to standardize on one analytics tool is a significant topic, but from my perspective it's simply important to consider what the landscape of Sitecore Analytics provides.  On the surface:
  1. Sitecore has incredible analytics drill-down capabilities--from a list of recent visits, to all visits from a particular company or source, to an individual visitor, to an individual visit.
  2. Sitecore offers standardized OLAP cube data, and a set of sophisticated business intelligence dashboards
  3. Probably most importantly, Sitecore provides ACTIONABLE analytics.  Your reports are right there in your power tool of content authoring and publishing.  Now you can DO something with that analysis right away.
  4. Sitecore provides incredibly compelling segmentation possibilities.
Together these capabilities span the spectrum of the ultra-specific (an almost Tealeaf style capability of watching an individual session which could be interesting for many situations such as a high-price ecommerce transaction, evaluation of a sample session for usability, etc.) to the ultra-summary (the rolling up and tearing back down based on my own business’ dimensions or facets).

So for today, let’s get ultra-specific.  For those of you that have read other articles about Launch Sitecore, you’ll know that our purpose is pretty specific—create an evaluation package that’s best-in-class so that our prospects can quickly learn that Sitecore is the best platform choice on the planet.  For those of you that understand our Engagement Plans, you’ll know that this process looks like this:
  1. Through demonstrations and technical deep dives, our Sales Engineering team builds interest in having a prospect visit www.launchsitecore.net.
  2. While enjoying some articles about Sitecore, our visitor is really urged to login or register to get to the good stuff (the download of the package itself).
  3. After registering, the visitor wants to download the package (the site itself that they can apply to their local test installation of Sitecore).
  4. After a glorious evaluation of the product, the visitor (now our customer) tells us how great the experience was in an incredibly over-simplified feedback form.
So, since this is only intended to be marginally real-world, let’s go through this is a visitor, all in one visit and see what we get on the other end from the reporting side.

The Simulation:

First, visit www.launchsitecore.net.



Since we’re simulating a brand new visitor, we’ll go ahead and register:

We are so excited about Launch Sitecore, we want to get our hands on the package right away, and we choose to download the version for Sitecore 7.x:



As we read a Digital Marketing System focused article, we check out the slide out at the top right of the screen, used to show the type of information Sitecore is collecting about our visit.  We notice that we've been categorized in the Digital Marketer Audience Segment (based on a number of articles we've read, including the "Create a Goal" article.  We also notice that it seems we are investigating the site to a level that shows we have a "Detailed" interest in the content.  We're not just reading the home page--we've investigated some detailed articles about the Sitecore Digital Marketing System.


The Reporting:

In the Sitecore Marketing Center, we can find the Latest Visit report.  Lots more to talk about around this reporting interface itself (including filters, date ranges, etc.), but for today let’s get to the info about our simulation.  First, since I’m in my home office on a snowy New England day, I’m behind my Comcast ISP account.  Beyond the scope of this article is my ability to tag my visitor sessions with additional information (from forms filled in, from authentication events, etc.) that could further identify my visitors.  In addition, I chose to “categorize” all Comcast Cable visitors as “My Company”.  (I could have categorized this as “ISP” or anything else I’d want to use to describe visitors from the Comcast DNS resolution).

I could drill down to all Visits from Comcast Cable by clicking on that link.  That would then show me more specific visitor sessions (possibly with specifically identifying tags).  Since I just ran through the simulation, I know this session is mine and I’ll bypass that by clicking on the link under the Date & Time.

image

Now at the Visit (Session) Detail report, I get a nice snapshot of my overall visit.  Some things to notice:
  • We were able to tag the session with “MikeBlogTest” (the name I used to Register with the site)
  • We see that I visited by clicking on a link within my Blog
  • We see the total value of the visit was 85 Engagement Value points (with the breakdown of goal achievement following)
  • We get a nice breakdown of the Pattern Card shape that is emerging from my visit—the specific content profile key attributes that are accumulating based on my content consumption.
image

Further down in the report, we start to get even more detail about the visit, including:
  • The fact that my Campaign was triggered
  • Fields that were filled out in our registration form
  • Pages visited and specific duration for those page views
  • Goals achieved (Register)
image

Another page, and more detail to round out the entire session.  More goals and fields filled in.  One interesting thing I want to point out here is that we actually captured an “error” on the site (line 17).  Error is in quotes here since, depending on how we handled it in the application, this could have been completely invisible to the visitor.  While this isn’t in any way a sophisticated error handling technique (it’s not meant to be), it could provide incredibly valuable information to the reviewer of the report in the context of the site visit.  What this tells the reviewer is that when a visitor clicked on a Search Result, there was no presentation assigned to the item clicked on.  A nice bit of usability testing after the fact (hey, we can’t catch everything in QA).

image

The Summary:

While not every analysis requirement will focus on an individual session or individual visitor, it is important to know that we can drill down to this level.  This gives us an incredible foundation to consider ties to transactional systems (Ecommerce), CRMs and other data sets to develop a full picture of our customer experience and value.  This is the ultimate leaf of the analytics tree, while keeping the ability to aggregate and roll back up to the branches.  By rolling back up, we will be able to get this same insight at the company level (all visits from one of our prospects), at the Pattern level (with Sitecore's ability to model the audience segments, now I can use those segments to hold aggregate analytics), and much more....