SES NY - Day One

The NYC version of the Search Engine Strategies 2008 Conference and Expo kicked off today at the Hilton on 54th Street and 6th Avenue. I missed the morning session due to work commitments, but made it in time for two sessions.

For the 2pm time-slot, I attended Mobile Local Search: Finding the Way. The session description offered "...best practices for answering needs of today's demanding [mobile] searchers," but I found the session to be largely a marketing jaunt for the companies represented by the four panelists. Nothing really to learn, with the exception that ~90% of callers have no idea what they pay for a 411 call.

At 3:45pm, the only option was Orion Panel: Getting Vertical Search Right, which proved to be more interesting. While, again, each of the panelists spent a fair amount of time working the PR angle there was a reasonable amount of substance to the conversation. A few highlights:

Steven Krien from OrganizedWisdom.com believes firmly that the future of vertical search lies in human assisted search, including an available service layer. As an example, Organized Wisdom (which focuses on the Healthcare vertical) delivers Mahalo-esqe human-curated search results. Also available, for $1.99 a minute, a site visitor looking for more specific information can talk (online) to a doctor. Putting aside the question of what type of doctors are willing to work for a start-up web site at a rate far below what they get paid at their day job, using real people to prepare search results for specific search terms is a strong value proposition. The example he gave was the Johns Hopkins Medical Center Forum on Pancreatic Cancer. According to the researchers at OrganizedWisdom, this forum is the best place to find an active and meaningful community of people dealing with this disease. Despite that, this forum ranks poorly from a natural search perspective and without human assisted search it would go largely unnoticed. Of course this gets to the relevancy question when trying to understand why these forums don't rank and it's not surprising that the folks attending Johns Hopkins Pancreatic Cancer forum have more important things to do than run out and create some back links. If you're not familiar with my views on relevancy, give them a read and let me know your thoughts.

Jason Finger from SeamlessWeb and Paul Forster from Indeed.com both mentioned how difficult it is for general search engines to utilize structured or meta data to help rank search results. The obvious reason for this is that the form of the data will vary from one vertical to the next. A common architecture utilizing an entity design pattern might enable a generic search engine to handle this from a purely technical standpoint, but it's simply not scalable / practical for a company to build the expertise in EVERYTHING that would be required to detail the data structure for all verticals.

It's interesting to watch this online search evolution. The failings of the generic engines shines a light on the need for vertical focus. This is clearly a case where "Jack of all trades, master of none" applies.