“Sears moved workload off of mainframes onto Hadoop and saved $2M per year,” said Aashish Chandra, DVP of Sears Holdings and GM of Metascale, which is Sears’ big data division. Chandra revealed this and much more at the recently completed RIS Cross-Channel Executive Summit, wihch ran from September 25 to 27 at the Four Seasons Dallas at Las Colinas.
Chandra also noted that the switch to commodity servers makes sense because “45% of the capabilities on a mainframe are never used and when you switch you can save 60% to 80% of your mainframe costs, which is important because IT budgets are not growing at the same pace as data volumes.”
Chandra’s session took a deep-dive look into big data, which is something rarely done at retail conferences. Big data is such a new and important field of expertise that most public speakers are consultants who talk about high-level, non-specific attributes from a 30,000 foot level.
Sears, on the other hand, is a true big data pioneer that has learned, made mmistakes and achieved success by hands-on effort. It currently operates the largest enterprise deployment of Hadoop, according to Chandra.
Want to publish your own articles on DistilINFO Publications?
Send us an email, we will get in touch with you.
Some key takeaways from the session include:
- Hadoop is inefficient for small datasets but exponentially better for large datasets compared to using traditional databases.
- Sears big data program is not only charged with solving analytics problems but also to be a revenue rainmaker, which it is successfully doing.
- Amazon used big data capabilities last year during Black Friday to seek out door-buster deals throughout retail and then change its prices every hour to undercut them.
Date: October 1, 2013