Predictive Models on Omniture/SiteCatalyst

[narayanank84409]

Scenario - I need to run statistical analytics on 1 year worth of data captured in Omniture. Was trying to understand what are the best possible ways to do it. Some of the ideas/thoughts I have are as below. Request to please share your thoughts

1. Integration points of SiteCatalyst with other BI tools for advanced/custom analytics.

2. How much data can it store at any given point of time to do active analytics. Can I do week/month/year comparisons and still be getting back data in decent time?

3. Is it advisable to take all this analysis by pulling data feeds from SiteCatalyst to a NoSQL DB like MongoDB or pulling the data into Hadoop for historical storage and run R/SAS analytics on them and/or using Visualization tools like Tableau etc. to run active analytics. 

Trying to find a scalable solution for problem scenarios like campaigns this year vs similar campaigns last year, predictive models, basket analysis etc and many more.

Thanks in advance.

[brian_au · Accepted solution]

I would highly recommend working with the full raw clickstream data or leveraging a solution like Data Workbench (part of Adobe Analytics Premium and formerly called Insight) to accomplish predictive modeling/complex statistical analysis.

Best,

Brian