Search has dominated content discovery and research for the last 10 years. However, the advertising and traffic dominance leveraged by Google’s search algorithm excellence has slowly been subsumed in the last 5 years by Facebook and a rise in social signals to drive traffic around the web. In 2009, Google reacted to this trend with a “personalization campaign,” a shotgun approach to algorithmically providing Search Engine Results Page (SERP) relevance by 57 signals around the user’s identity, rather than simply semantic language queries. There is no “standard Google” any more. Eli Pariser’s recent work indicts this trend, and suggests problems with information retrieval in an algorithmically-personalized filter bubble. There are implications for search engines, for users, and for brands; but I want to focus on the implications for anyone who creates and curates content for online distribution. Journalists, editors, and content strategists.
For those of us that create and curate content, search is something we recognize as a foundational commodity. Our job is generally cultivating a long-term relationship with the customers/audiences, punctuated by elements of transactional marketing. The process as I’ve generally participated in it involves SEO as an integral part of the planning process, but deprioritized in favor of clickstream and referral data as the pages mature. This is an “exploratory search” perspective, so it’s helpful to provide some historical context before talking about the impact of some of the recent social shifts in search results.
There are a couple of different ways to talk about search as it’s evolved over the years. These are really organized by task and query depth. The task is really “what do you want to know?” In the early days of the web, the type of answer you wanted determined not just the type of query but also the engine. “Exploratory search” really answers the “what do you want to know” question with “I don’t know – just give me everything on this topic.” This is generally unstructured data (lists of results organized only by relevance) in the Search Engine Results Pages (SERPs).
Faceted search, is a more recent entry into the online search space. Rather than offering the black box experience of a Google or Bing front page, users can filter and pivot the query, personalizing the relevance. A good example is Yelp, which focuses its search results around user-queried local business. There are filters for business types, rating, neighborhood-specific proximity. Parallel to this rise is the growth of folksonomies, do-it-yourself classification strategies that are multi-topical and not necessarily hierarchical. Witness the power of Delicious – a social bookmarking site – that allows you to discover other user’s relevant bookmarks through culturally common “tags.” Click on one of those tags in my tag cloud, and then look at the tags in the right nav that are related to that tag. This can be a method for both filtering and discovery of new topics around a given concept.
In the examples of faceted search above there’s a heavy leveraging of social media, which is mostly unstructured data. But faceted search can also be a wizard-oriented approach to structured data. The rise of mobile and its virtualized search box without regard for the computer desk has resulted in narrowly focused task- and location-based faceted querying. Queries are action-oriented and one answer in scope. An example is “Seattle Weather.” You don’t need 15 pages of results, simply 1 page of authoritative sites with accurate forecasts. Bing’s marketing campaign for the last year or so has attempted to differentiate its results offering as a “Decision Engine.” Another example is CNet’s Cell Phone Finder.
What’s the important editorial takeaway for this differentiation? Search intention is different than it was 5 years ago, and information architects need to bear in mind the importance and implications of search as a foundational commodity. We once only wanted to discover. Now we also want to accomplish. What intention does your content surface? Is your content transactional and static? If so, traffic will be from well-defined sources (company home page, exploratory search, etc.), and keyword strategy becomes key. For content that’s technically deep, relying on multiple pivots to slice and dice particular topics, faceted searches become a lot more important to discovery, and social media along with strategic linking strategies will prove more impactful to customers. In the next part of this series, I’ll focus a little more on influentials and curation as a new signal for search engines (and enterprise search) to incorporate.