Last week kicked off Hadooponomics, a big data trends podcast produced in partnership with Blue Hill principal analyst James Haight. Over the next four weeks, we’ll be bringing you a new episode every Wednesday.
This week, James sat down with Dr. Claudia Perlich, chief scientist at Dstillery, and Shant Hovsepian‘, CTO and co-founder at Arcadia Data, to discuss the progression of data science in the ever-changing digital advertising world and how big data plays a role.
Looking at the hypercompetitive world of advertising, Claudia touches on how big data helps companies like hers enable hundred-millisecond decisions about which tailored message to share with which person on what device, 50 billion times a day. Thanks to machine learning, analysis of individual user behavior instead of serving ads based on demographics is re-writing the rules of the advertising market. Big Data 1, Don Draper 0.
Shant nicely complements Claudia’s mass customization offering insight on the how value density of this data is changing. A lot of big data comes from recording everything. The thing is, it’s useless unless you intersect it with other data attributes and contextualize it. Doing this efficiently means you need an architecture that does two critical things:
- Makes it easier and faster for people to get the type of analysis that they want across a wide swath of different data points
- Puts the data in an operating platform that doesn’t break the bank as data grows or blow your mind with endless management complexity.
Check out the rest of the podcast here. Stay tuned for next week, when James will be hosting Eddie Garcia, Chief Security Architect at Cloudera, just in time for Strata.