Lots of data is great, but knowing how to get value is better


So you’ve got a lot of structured, semi-structured and unstructured data. How do you or your business get the most value out of it? Nick Weir, CEO of ChoozOn, which uses intelligent matching to connect people with the best deals, gave some basic tips on the most important things to consider when mining your data in a presentation here Wednesday at Structure:Data 2012 in New York City.

Nick Weir at Structure:Data 2012

(c) 2012 Pinar Ozger. pinar@pinarozger.com

As the former vice president of data strategy at Yahoo(s YHOO), Weir has plenty of history with poking around large amounts of data. From that experience, the biggest challenges, he says, are: being able to truly exploit all the data available to you, proliferating analytics throughout your organization and driving significant business value from it.

The data he gave as an example: the 800 million Google(s GOOG) queries per day, 250 million tweets per day and 800 million Facebook updates per day that shows what customers are either searching for and what they’re saying about your brand online.

The best way to squeeze actual value out of all these tweets, updates and searches, he said, is:

  • Understanding user intent
  • Knowing how people generally feel about your brand
  • Pinpointing context: where and when people are talking about or searching for your brand

He gave an example of what Yahoo has been able to do. They take their gobs of data they suck in every day from their millions of users and do behavioral targeting to get higher value CPMs, Weir said. “Yahoo constructs DNA for each user. They construct models that look at behavior, score users on a number of dimensions and target ads to those users,” he said.

Besides crunching all that data, there’s a lot of modeling involved, he noted. “Not just the intensity with which [someone] is looking at a topic, but recency matters” too, another example of how important context is for deriving value from having a lot of data.

Watch the livestream of Structure:Data here.

gigaombigdata on livestream.com. Broadcast Live Free

Comments are closed.