PREPARED FOR
2007 IA Summit: Las Vegas
SEO and IA: The Makings of a Beautiful Friendship
ASCENTIUM 225 108th Ave NE, Ste 225 Bellevue, WA 98004 t 425.519.7700 f 425.519.7758 ascentium.com
Introduction
Me: Information architect/Search specialist IA since 1998 Search since 2004
Topic: Search engine optimization and IA Shift in user locus of attention From navigation to search box Shift in our locus of attention From macro-structure to micro-wayfinding What I want IA to become a partner in developing search technology that works with the user IA community to “think” about how users find their websites when they design them Key takeaways [fingers crossed at my end] Search optimization and IA can and should co-exist One should not exist at the expense of the other
ascentium.com
Search Usability
Web analytics show preference for search box over any site navigation of any kind Search enables users to develop a need-specific/use-specific information path Search engine users visit more pages than those using navigation Pogo effect Ask.com now offers preview service so user does not have to click through How much of the navigation will they see in a thumbnail?
Out of the top 20 results and you are out of sight and out of mind for a majority of users
ascentium.com
Blame it all on Google
PageRank is a pre-query valuation Based on number of links to the page 1 link=1 vote Most votes wins top placement Has no relationship to the subject of the query Googlearchy : dominant Web sites become more firmly entrenched in search results by nature of size Link rich get richer Failings soon uncovered Link farms Googlehacks
ascentium.com
Search 2.0
Web 2.0 give us Search 2.0 Harnessing the collective intelligence Online bookmarking Architecture of participation Open source Search Peer-to-peer Search Index of nodes in system Query passed to find appropriate node Remixable data sources and data transformation Local search Any of the “maps” applications Kayak.com and other travel sites Software above the level of a single device Mobile search Compensation for the commercialization of organic search Paid ads do not have to map semantically to the results they accompany Wales and Searchipedia Program not tied to a revenue model
ascentium.com
Now It is All About Meaning
As Moore’s Law brings about cheaper, faster, stronger hardware, the quest changes from indexing everything to the presentation of results Search challenge to determine relevance without understanding meaning
Transition from strict computation to computational techniques to determine meaning Hilltop Algorithm Topic-sensitive PageRank
ascentium.com
Hilltop Algorithm
Segmentation of corpus into broad topics Subset that is then extrapolated to Web as a whole Created by Jon Kleinberg at Cornell in late 1990s Consultant to Google Selection of authority sources within these topic areas Authorities have lots of non-related pages on the same subject pointing to them Quality of links more important than quantity of links
Determination of HUBS Pages that point to many authority sources
Pre query calculations applied at query time Likely part of Google’s Florida update in 2004
ascentium.com
Topic-Sensitive PageRank
Consolidation of Hypertext Induced Topic Selection [HITS] and PageRank Pre-query calculation of factors based on subset of corpus Context of term use in document Context of term use in history of queries Context of term use by user submitting query Creator now a Senior Engineer at Google
ascentium.com
Search Further Down the Road
Semantic search technology patents
Search tool with preset categories and keywords
4-part database of information Index, categories, keywords, document-specific data Categories define topics through human-mediation Keywords extracted from document text User can iterate search results through related keywords presented from database Brokering application that facilitates selection of best search engine for the user’s query Creates “sketch” or compact representation Compares sketches based on determined similarity threshold Deleted duplicate entries
Search manager
Similarity estimation
Personalized search
Microsoft: Compares snippets of Web search engine results with data collected from user behavior and client Demonstrated in NYT article March 7, 2007 Google: user bookmarks [online and client] used to construct “personalized search object” that is used to filter Web search result
ascentium.com
Predictive search
Bayesian model Compares user choices to predict more appropriate result from same vector space
SEO and IA: Choices
Capitulate No action Search technology continue on parallel path Cooperate Work with current search technology Develop best practices that build on developments in search technology Initiate Influence development of search technology Become a partner in developing user-centric search technology
Action Items
Influence the technology to work for not against user
ascentium.com
Site Navigation Strategy Site Organization Strategy Link Strategy Page Code Strategy Content Strategy Metadata Strategy
Initiate: Site Navigation Strategy
Locus of attention has changed from navigation to search Hard-coded navigation structures are losing ground to pogo strikes
Navigation Blindness Navigation Fatigue Page Paradigm Transitional Volatility
Users need inducement to move further into the site Search technology rewards relationship navigation Berrypicking Information Model
System approach to navigation development Systems have specific behaviors and outcomes
ascentium.com
Initiate: Site Organization Strategy
Distance makes a difference Hierarchy reflects relevance MOSS 2007 and use of structural factors
URL depth: the further from the homepage, the less important it must be Click distance: the further from an authority page, the less important it must be
Architecture extends from the site to the page
ascentium.com
Initiate: Linking Strategy
Links are human-mediated relationships Blast services are no longer worthwhile Related sites, niche directories, online bookmarking sites, provide starting points Create link-based relationship model of relevance Create or find authority Hook up to HUBs Think beyond the site
ascentium.com
Cooperate: Page Code Strategy
Reveal the site to the search technology Sitemap.xml Provide on the page navigation Don’t rely on dynamic navigation that spider cannot read
Craft structures that cue technology on importance
Illuminate the non-textual functionality Optimize JScript and Flash
ascentium.com
Initiate: Content Strategy
Dense, subject-specific content is what is indexed People will scroll If they don't scroll, they will print it out
Content to code ratio of 25%
Promote a keyword-to-content ratio 10–15% Design on-the-page structure to move important information to the top Design relational content models Next steps as well as more information Develop authority sections on site Topic-based, not type-based
ascentium.com
Cooperate: Metadata Strategy
Many forms of description In the code Page title [in the browser window] Description Keywords? In the content Display title Content headings Most effective if unique to the content on the page Say goodbye to cut and paste
Description rivals structure for importance for user context Ask.com thumbnails
Humans determine the “meaning” of the document and inform the machine
ascentium.com
SEO and IA: Threats and Opportunities Threats
Opportunities
Search technology advances without user representation Search engines have become dominant navigation tool through information spaces Bountiful Relevant? Traditional IA methodology increasingly less useful Hierarchy: pages further from the home page deemed less important Hard-coded navigation: not visible to search engines Not Authority-based
Users seeking human-mediated guides to find information Current search rewards a more flexible and intuitive IA Replaced by a new structural paradigm based on relationship and context Hub and authorities Quality over quantity Birds of a feather subject-wise
ascentium.com
Marianne Sweeny marianne.sweeny@ascentium.com
ascentium.com