In "What History Tells us About Facebook's Potential as a Search Engine, Part 1," we explored whether Facebook, which hosts more than 600 million U.S. searches each month, had the potential to become a serious player in the search market.
To effectively evaluate this potential, we must first get a better understanding of the ways in which consumers are searching today. One particularly telling search metric we like to evaluate is the average number of words per search. This metric, examined over time, offers an indication of how consumers are adapting their search behavior to shifts in technology and culture.
Words Per Search Increasing
The first thing we will notice in the chart below is that the average number of words per search at the top search properties has been steadily increasing over the past two years to approximately 3.2 words per search. This trend suggests that consumer search behavior is evolving toward a higher level of specificity, and is a far cry from the days when consumers typically searched on a single keyword.
Searches on Facebook, however, are quite different than those on the top engines. For starters, the average Facebook search contains only two words. That trend hasn't increased in any meaningful way over the past two years.
Perhaps these numbers shouldn't be too surprising given that the majority of searches on Facebook are searches for people -- first and last name. But delving a bit deeper into the list of top searches on Facebook, beyond the "first name, last name" searches, people are beginning to search on more traditional search topics, things like "games," "shoes," and "iPad."
Non-people-based searches on Facebook look somewhat similar to the types of searching we saw during the early evolution of the search engines.
If these are the same people who are typing three-plus word searches on Google and Bing, why are they still relying on such rudimentary search terms on Facebook? When people are unfamiliar with a search experience, they prefer to dip their toe into the pool rather than dive in head first. Until they get a better sense of how the engine will respond to their requests, they keep it simple.
But the fact that we're seeing the first real signs of a burgeoning "traditional" search experience bodes well for the future potential of Facebook as a search engine. We should see this type of consumer behavior evolve along the same lines of traditional search as more dollars flow towards social media.
Social Search Filter
It will also require further development of the search technology on the part of Facebook. The true promise of traditional search on a social platform like Facebook is the ability to apply a social filter to search results.
When I search for "shoes," I want to see which of my friends are fans of Nike, or which ones "liked" that pair of Pumas being sold on Amazon, or which ones raved on their newsfeed about their comfortable new running shoes. For now there are some options allowing for broad matches to Friend & Everyone posts, but much more can be done.
Facebook must improve their ability to relate your search results, on the first page, to you as an individual. Search engine algorithms do their best to deliver relevant results based on all of the people who searched on similar terms as you, but Facebook has the unique position with which to analyze your personal relationships and further differentiate results.
The Facebook display advertising targeting features are quite impressive, considering factors such as your latest updates, your "likes" and those of your friends, but its ascension as a search property will be directly impacted by its ability to do this same sort of qualitative analysis of you and your entire circle and customizing the search results accordingly. Being that they partner with Bing to deliver the web search results, it puts Bing in a strong position from which to innovate using an incredibly large and yet closely personal data set. The ideal Facebook search result page format would be universal in nature, with a collection of the most relevant internal Facebook information combined with appropriately filtered external web results.
As consumers become more comfortable with an inherently social search experience, their search behavior is likely to evolve with it. We will see the average length of their search phrases increase, and it will be driven by both technology innovation and more relevant marketer content before these consumers make the jump.
We're still early in the evolution of social search, but the nascent search behaviors we see developing on Facebook suggests it not only has the potential to become a viable search engine, but in fact it has a chance to help redefine the way we think of search. The social filter can provide a new layer of meaning to search results, but it will be a matter of technology, marketing, and consumer behavior evolving until we get a true sense of what the future of social search will bring.