Search Engine Strategies Toronto 2010 started with a bang. The conference is being held at the Hyatt on King Street west which is proving to be an excellent venue. On the smaller side, it is providing the optimal density for networking and 'vibe' and with 1100 or so participants every session is full and the vendor area always has a buzz to it.
Keynote speaker, author and the 'grandfather' of the discipline of Information Architecture, Peter Morville was insightful and entertaining and provided an interesting look at search and information architecture from a few different angles. His entire presentation was like a survey of different innovations in designing for search.
He discussed search behavior as a 'berry picking' exercise rather than a 'search and destroy' exercise; characterizing it as a learning process with the user taking a little from the first site they visit in a search, modifying their query with what they have learned, searching again and so on rather than a search, find, consume pattern.
Another point of his session discussed the relationship between searching (finding) and browsing (discovery) and how each can be more appropriate in different circumstances. When options and complexity are low, best practices typically start with browse behavior and then transitions to search behavior as the level of complexity grows. For example, interfaces for online television started with browsing which was successful until the number of available channels skyrocketed at which time a combined search / browse model became more appropriate.
Without a doubt the most interesting part of his discussion was his look into faceted search and how some are succeeding at it (think shopping.com) and others are struggling, the most prominent being Google's well published experimentations. Faceted search is the concept of aggregating data from different sources but providing a unified search, browse filter system over top most commonly seen with online ecommerce aggregators.
One of the other trends in designing for search he discussed was the transition to actionable results tied to search, particularly in the mobile space and the development of a search/action interfaces; think find/call or find/map or find/play(music) models.
He then transitioned into wondering where the future of search was headed and gave several hints:
- Question Answering: WolframAlpha a computational 'search' engine
- Decision Making: Hunch a 'learning' recommendation engine
- Understanding: Oakland Crime Spotting a GIS based systems
- Multiple Sensory Search: the way out future: we already have find by sound Shazam or Midomi so think find by smell (think recipe site) by touch, etc.
Overall, Peter's presentation provided a good look at the world of search from a design perspective and put some theory and understanding behind many of the innovations that we take for granted. He is adept at showing just how different these design changes are and when they are viewed and discussed in succession it becomes obvious what real innovations they are. I can see picking up a copy of books in the near future!