In the first part of this two-part post on Digital Analytix, Chris Meares and Royce Fung reviewed the solution’s segmentation, database architecture and reporting UI.
In this follow up, Ryan Praskievicz takes a look at another area where Digital Analytics appears to trump SiteCatalyst…cardinality. If you’re not familiar with the term cardinality and you use SiteCatalyst, perhaps you’re familiar with the term “uniques exceeded.” Ryan explains how this affects your web analytics reporting. I’ll finish up with a short section on pricing.
Ryan Praskievicz @ryanpraski
Cardinality Comparison – Digital Analytix, Google Analytics, Adobe Omniture SiteCatalyst
Digital Analytix (DA)- has no variable cardinality limitations. If you have millions of unique products, campaign identifiers, or any other data with cardinality issues you can think of, DA can report and segment on all the values. It will be interesting to see just how fast DA returns reporting queries on multiple dimensions and metrics with millions, even billions of unique values.
Google Analytics (GA): there is a 50,000 daily unique value limit for each table (a table being a dimension within GA like Page, Event Category, Event Action, Event Label,etc). Once a table hits the 50,000 limit all the subsequent values are grouped together as (other). Once captured there is no way to know way to see what is in the (other) group. There are possible workarounds like creating multiple profiles to capture different pieces of your data. For example if your site sells electronics you could create separate profiles for Televisions, Computers, Video Games, etc. Then you would be able to see 50,000 unique daily products for each of the profiles. This would ensure that all your product data was captured, but the drawback of this set up is that your data is sits in silos by product type and analysis across profiles is not possible.
Google Analytics (other) for the pages report
SiteCatalyst: there is a 500,000 monthly unique value limit per variable. This applies to all SiteCatalyst variables including page name, campaign tracking code, products, referrers, etc. Once the 500,000 unique value limit is hit, all subsequent values for the month are grouped together as Uniques Exceeded. Where Omniture differs from Google Analytics, is that even though the SiteCatalyst data is being bucketed as Uniques Exceeded, the contents of the Uniques Exceeded bucket can still be accessed. To get reporting with no cardinality limits in Omniture you can use DataWarehouse. The drawback of DataWarehouse is that for queries with large amounts of data may take hours or even days to return results.
SiteCatalyst Uniques Exceeded
Discover: there is a 500,000 monthly unique value limit per variable. In Discover the same cardinality limits as SiteCatalyst apply. In Discover each unique line item beyond the 500,000 monthly unique limit is shown as Unknown rather than in SiteCatalyst where the data is grouped together as a single line item: Uniques Exceeded. Again you always have the option of using DataWarehouse to get actual values for the Unknown line items in Discover.
While I can’t discuss pricing in specific terms, based on what was presented, the model is based on server calls as is the norm. However, Digital Analytix uses a two cookie approach to process data. This sends data into a SuperSite (same as a Global Report Suite in SiteCatalyst) and into subsidiary report suites. This eliminates the need for multiple server calls and consequently potentially less expense. It does appear that there will be modules that while optional, will likely be necessary if you want to have a robust and complete measurement solution. As with any fee-based solution, I recommend that you understand what you’re paying for and how what’s featured in the demo translates to your contract so there are no surprises.