Peter is a librarian (!) turned information architect.
“The single biggest opportunity to make positive change is through search.” The return on investment can be significant – higher sales, lowered reliance on customer support.
A search system includes content creators, metadata and indexing, tools (algorithms), interfaces, and end-user psychology and behaviour.
In mobile search, we nudge people towards browsing – many mobile applications are so new that people don’t know what to search for. We have to help them understand what’s possible.
Search is a hybrid of design, engineering, and marketing. There are complex interdependencies and changing requirements. It’s a problme that is never solved. It’s a project and a process that never ends.
Beyond usable: accessible, useful, desirable, findable, credible, valuable.
Morville created the search patterns library in Flickr. Patterns include narrowing, expanding, thrashing (when people modify their searches over and over again, trying to get good results).
Auto-complete: why wait for the results interface to start helping users? Suggest options as soon as they start typing.
“Best first” is the most important search pattern – the most relevant results must be at the top.
Speed is critical – that’s why Google has invested in performance and still displays the time it took to do your search at top right. Search must be fast.
Social data helps inform relevance ranking. (Our number of copies ranking works like this. Example of Books & Authors, where this doesn’t work.)
Mentions challenge for libraries of exposing licensed databases. How do we design an experience that doesn’t rely too much on where the data comes from?
Faceted search: facets serve as a custom map to search results that helps users understand what they’ve found. It also clarifies the options for next steps.
NCSU is using Summon in beta to integrate federated search of their databases.
Amazon iPhone app – good example of faceted searching presented in a mobile interface.
Amazon gives you limited facets with initial search results, richer facets once you begin to refine.
Attention to detail and continuous incremental improvement contribute to a great search experience.
iPhone apps already let you search by sound (song recognition) or search for things that look like an image. Input from users does not have to be verbal.