FT's Newssift: Power of Metadata Exposed (Finally!)

There is so much to celebrate with the Financial Time Group's debut of Newssift. Like a debutante at her coming out ball, Newssift turned heads and received appreciative glances. The question is will the market step up and marry itself to the new business search engine (a flirty glance is one thing, parting with actual dollars is a real commitment). The site, which is currently in beta is free -- for now. The goal of FT is to get visitors to consider the site their home for business intelligence -- after parlaying subscription dollars.

The real beauty of Newssift is in what you can't see: The metadata generated by semantic analysis. Yes, the interface offers visitors a very simplified mechanism from which to find the information they need. But it doesn't matter how sexy the interface or powerful the search engine (in this case Endeca) if the metadata extracted is not linguistically and computationally sound. As the online media expert for Nstein Technologies, I've been aware of this project for some time -- and am thrilled with Nstein's role in generating this rich metadata -- and am further thrilled that my job of educating others to the value of semantic search - is being made so much easier with Newssift's launch!

Unlike most search engines, which basically start with a blank slate and require users to type in a phrase, Newssift anticipates that a reader might be interested -- in what others are interested in. A data table presents top news of the day -- broken down by People, Places, Organizations, Business Topics & Themes. Readers can select one of those parameters -- or use the search bar to type in their own queries.



Now here is what makes Newssift unique. Like the name suggests, readers can continuously sift through information until they find the pearls of wisdom they are looking for. Each time a result is displayed -- readers can refine their search (mine down) using the table or the search bar. This may be one of the best instances of having these two querying mechanisms coupled to work together - because both are searching the metadata! (Too often the two methods are exclusive of each other.)

The result is that the interface provides a type of funnel to the information desired. The constant changing of the column parameters allows people to see what other information may exist -- enticing them to read things they may not have even considered. This process is critical information publishers for several reasons:

1. The process itself means people will likely stay longer and consume more pages. (Robin Johnson, CEO of Newssift, told me that early metrics showed a 3x increase in page stickiness -- all through contextual linking).
2. People don't know how to find information -- because they don't know what question to ask! (A fine list of why search engines are typically broken here.)
   
3. Seredipity is brought back into the equation. I may want to find out what's new about Swine Flu, and when I see Italy as an option -- it may entice me to discover what the Italian slant might be.
   

This initiative was in the works for nearly two years. And if you have ever hatched a brand new business within the walls of an old stalwart one, you know it is no small task. Johnson and his team not only had to create a business model that worked -- but they had to assemble a team of vendors like Nstein, Endeca and Lexalytics who could see the vision, adapt to the clients' needs and make the whole thing crystalize.

Newssift has accomplished much just by just package news in a faceted format that will entice many to dig deeper and in backing up this form with semantic metadata to make sure that the most relevant information is provided. I don't have a crystal ball, but my sense is Newssift will become a shining example of how information can be found, explored and digested -- putting to rest forever the idea of mere search.