What is the Data Telling us?

Marketwatch's Bambi Francisco moderates a panel on "What the Data is telling us?" at the Stanford Summit today. She talks to Chief Data Officer and SVP of Research at Yahoo Usama Fayyad, Ask.com's CEO Jim Lazone, Peter Norvig, Director of Research at Google and Quigo's CEO Michael Yavonditte.

Bambi asks each of the panelists: "Give me three observations about how consumers behave and what they want."

The notion of the portal is still very relevant, and the data behind the usage on the Yahoo front page completely supports this notion. Says Usama Fayyad, "Adding a news preview box in the mail. When people are doing email, they're also looking at news. Search is not enough. Tagging based sites like flickr is another indication can help make information more relevant.

Jim Lanzone of Ask.com says, "we find that most people are lazy and may not tag. We have come up with products that suggest searches for them to help people narrow data down even faster. We learn that on Monday afternoons, people will do the tasks that they meant to do on the weekend. On the query stream, we learn that Tuesday people are already looking forward to their next weekend.

Peter Norvig, Director of Research at Google says, "For language, when I was an undergraduate at Brown University, there was a billion times less access to words and language than we have today." He goes into how they can now learn about specific facts because of data that was not available before.

People agree that the search interface is broken. Jim says that users make a trade-off, meaning that they decide that its not worth their time to toggle and figure it out, so they make a query and go, hoping that their search will get narrowed down. They don't have the patience to 'figure it out.'

Speech interfaces are one solution, but so is learning more about users, how they behave and how they make queries.

It's not one box and it needs to be different products for different audience segments. For example, its no surprise that Yahoo would learn that a younger population searches for more entertainment and music than an older population who may do more searches for Finance, childcare and healthcare.

What about data that shows how people are 'socializing?' Peter talks about lessons learned and how there are some things we just haven't been able to figure out. "We still don't know and we need more examples. For example, why has Open Directory stagnated and Wikipedia taken off?

Usama asks questions back to the panel and to the audience, suggesting that we're still learning. What is it that makes communities grow or shrink? What are the economic incentive models that make sense? What is it that is really going to build trust in the long term? A lot of these funamental questions, including user interface and search queries are begging for a deeper understanding.

Tag: AlwaysOn
Tag: Stanford Summit
Tag: AlwaysOn 2006