The search leader recently rolled out a wave of new products designed to provide consumers with a more personalized search experience. Unfortunately for Google, its quest to know and understand its users misses the mark.
As it stands, Google processes each search query and limits the relevance of the results based solely on the user's personal search history. The shortcomings of this approach provide an opportunity to look closer at personalization and define what steps must be taken in order for users to feel a search engine actually understands their interests and their intent.
When we type a word into a search field, we want to believe our favorite search engine intuitively knows what we're looking for. While currently a pipe dream, there are processes that will get us closer to this goal. Personalized search can prove far more reliable through an analysis of the individual's preferences and the characteristics that comprise a Web site. Only those search providers that understand these two separate qualifiers will deliver personally relevant search results every time.
To fully understand an individual's preferences, search engines must move beyond the simple keyword and focus instead on capturing the consumer's intent at that moment. True personalized search takes into account behaviors that are consistent over time--thrifty vs. luxury-inclined, or adventurous vs. cautious--as well as those that vary based on context, like shopping for a new computer, researching a project or planning a vacation. Combining recent online behavior with individual profile history gives the search engine context for understanding an individual's true preferences and intentions at the time of the search. Persistent behaviors are given more weight until the consumer makes it clear that fleeting interests should supersede.
Equal focus must be given to each Web site as well. Knowing whether a site is a blog, news outlet, e-retail site or forum helps pinpoint search results relevant to the user. To achieve this understanding, Web pages must be classified by subject (autos, personal finance, home and garden), as well as by content type: news, pay-per-download or rating and reviews. In addition, some sites cater to a particular lifestyle, an attribute that cannot be ignored when delivering relevant search results based on an individual's preferences.
Currently, Google employs a disambiguation strategy to help understand a user's intent. A query for the word "jaguar," for example, will prompt Google to use personal history to determine whether it's a search for a luxury automobile, football team or jungle cat. It's worth noting that a disambiguation strategy provides only limited value, as search parameters can change by the moment and there is no guarantee that results will prove more personal and more relevant.
With personalized search results based on the dual knowledge of the Web site and the consumer's behavior come sponsored links and targeted ads that are ranked by relevancy as well. This relationship provides enormous value to the consumer, as well as to the businesses paying big money for keywords and sponsored links. Personalized search fine-tunes the results so each consumer is highly qualified when he or she clicks on a particular ad or promotion. This approach also allows consumers to find the information they're seeking easily while building loyalty for the search engine that truly "knows the individual."
Until search engines embrace the importance of understanding both consumer preferences and the content they seek, the promise of personalized search will remain unfulfilled. It's time we give more to today's busy consumers. They expect that their trusted search engine knows their intentions beyond a simple keyword typed into a search box.
The search engine that gets personal and helps consumers find relevant content each time will lead the pack and stand the test of time.
Steve Johnson is chief executive officer of ChoiceStream, a provider of personalization solutions in Cambridge, Mass.