Search, while dominant, is not very good
Jim Kennedy is VP of Strategy for The Associated Press. He is also a member of the iFOCOS Search Working Group, which is holding its first meeting April 24, 2007, and of the iFOCOS Advisory Board.
Over the past six years, the online function of search has become the dominant content entry point for digital information consumers. More than 7 billion searches were launched in March, up 14 percent from a year ago, and the titans of online search – Google, Yahoo, MSN, AOL – touch nearly all of the active Internet users in the United States in the course of an average month.
Desktop search, by anyone’s measure, is the most popular on-ramp to the information superhighway. Social online networking may have gained some ground as an Internet entry point in the past couple of years – thanks to the MySpace-Facebook-YouTube phenomenon — but search has even greater days ahead with the mobile environment and the living room still up for grabs.
Think about it this way: While digital citizens may always need their “friends,” they’ll need a “remote” even more in order to navigate the many screens in their digital lives. Search is that remote.
If that’s the future we expect, then those of us in the information business – content providers and technologists alike – have a lot of work to do. Search, while dominant, is not very good. It seemed to plateau once it reached “good enough” for general use, and many stakeholders turned their attention to monetizing the traffic flow through advertising technology. That was sorely needed, of course, for revenue generation, but it didn’t do much to improve the relevance or timeliness of underlying search results.
The research firm, Outsell Inc., reports that the search “failure rate” for desktop information seekers keeps rising, and at last report was 30 percent, based on the firm’s annual survey. That suggests that search only satisfies the user seven out of 10 times, which really isn’t good enough for a dominant online entry point.
Before the world turned its attention to search and contextual advertising, search was improving rapidly. Indeed, Alta Vista and, later, Google rose to prominence on the strength of their information search capabilities in those early days. That led others to go to work on improving search, and enterprise engines, such as Autonomy and FAST, and many vertical search engines emerged and advanced search technology considerably.
Nonetheless, general online search still flunks 30 percent of the time. In the case of news, the failure rate could be even higher. Try putting anything close to a real-time keyword into the plain-vanilla search of the giant engines and see what you get. If it’s a running story, maybe you’ll get a development from a day or two ago, from a site that’s optimized for search but not necessarily an authority on the subject or close geographically to the action. At times, the results can be laughable.
That’s probably OK for the desktop, and we put up with it, because the next search is just a keyword away. But try that on a mobile connection, or imagine fumbling with a TV-remote search, only to get nonsense results. No one will be able to monetize that experience, and users likely won’t put up with it.
In fact, search might begin to fade as the great engine of digital revenue growth and give way to recommendation engines, or we might even fall back to a destination-based experience of some sort to minimize the pain of lousy search.
Perhaps these outcomes are overstated, but there’s no hype in the nagging failure rate of general search. What’s more, this failure rate is working to the distinct disadvantage of the very content providers who could provide the information consumers are looking for. Online newspapers, for instance, are struggling to surface in general search results, largely because their content is not optimized for real-time discovery.
How can this picture be changed in a competitive online world where content providers and distributors are focused on their own proprietary agendas? It would seem an impossible task. But what if the various “stakeholders” in search – content creators, technology companies, search engines and other online distributors – came together to identify key initiatives, which when combined might raise the tide for all ships?
That is the hope that inspires the working group discussion of search being sponsored by iFocos.org. As the organization has done for the citizens’ media space, with its We Media research and conferences, the objective of the working group discussion is to raise questions and awareness that might enhance the evolution of the search ecosystem.
That’s buzz-speak for a simple goal: Let’s open the floor for ideas that could improve general search. Are there initiatives various constituents could undertake – individually or collectively — that would significantly improve the search user experience? What are the costs and benefits of those initiatives? What is the potential return on investment?
For The Associated Press, I can speak directly to a major metadata initiative that we hope will establish some standards for the categorization and tagging of content across the daily news industry. We expect this effort to result in an AP “stylebook” for indexing news that would be roughly comparable to the writing and editing stylebook followed by news organizations around the world.
Standardization – in this case, for tagging – is one example of an initiative undertaken by one stakeholder that may be very useful to others. Are there more? Let’s talk about it on April 24.
