Future of Search: A Different Perspective, The

Design Management Review, Winter 2009 by Anderson, Stephen P

The mission at Viewzi is to dramatically improve the search experience. But, as Stephen Anderson explains, that doesn't mean developing a better search engine; rather, it means developing a better way to view search results. The results are a series of custom-tailored scenes whose look and feel change depending on the intent of the search. It's an impressive innovation that promises to reshape the Internet landscape.

Google recently celebrated its tenth birthday, sharing with the world a nostalgic recreation of the 2001 version of its site. Surprisingly, however, although there has certainly been a steady stream of subtle improvements over the years, the interface for today's Google search engine results page (SERP) isn't all that different from what we saw nearly a decade ago (Figure 1 on next page). This isn't a bad thing: Google, Yahoo!, and hundreds of other search companies have invested billions of dollars to improve the algorithms that spider the Web, index pages, and bring us back the information we want. And newer entrants like Powerset and Swingly are perfecting the math behind natural language recognition. Good search results depend on the engines behind them. But as the quality of results gets ever better, are we reaching a point where value comes from more than just good search results?

Author/professor Don Norman once stated that "when technology delivers basic needs, user experience dominates." The mobile phone industry is a great example of this evolution. No one would now consider entering that mar- ket with the original Motorola DynaTAC. What made it successful at the time - a phone that doesn't need wires - is now an expectation. Mobile phones have continued to become more reliable and usable and are packed with some useful (and not so useful) fea- tures. The introduction of the iPhone (and its imitators) raised the bar once again - not in terms of more features, but in the way in which people experi- ence information. Now our mobile phones aren't just took - they're also fun to use! This process of product maturity forms the basis for my User Experience Hierarchy of Needs model (Figure 2 on next page), in which I pro- pose that most technology product and service experiences go through six levels of maturity, from "Hey, this thing actually works!" to "This is meaningful in my life."

So where does the search industry fall along these lines of product maturity? If you look at many of the top trends behind search, much of the focus is still on how we procure results. Mahalo asks users to submit data. Rollyo allows individuals to create their own custom search engines, with results limited to selected websites. Similarly, Lijit is a site-specific search engine (installed on your blog) that can also search your wider social network. And then there are all the dollars invested in natural-language search. These are all worthwhile ventures; capturing even one percent of the search market can make you a billion-dollar company. But which do we need: even better results, or a way to make sense of the results we're given?

This is the question we asked ourselves when founding Viewzi - a new kind of search company. Instead of focusing on the search engine (we leave that work to Google, Yahoo, Microsoft, and other partners who excel at this), we focus on how people experience information. In short, we focus on search intent.

Framing the problem

As designers, we talk about the importance of properly framing a problem. For the last decade, the search problem has been framed as, "How can we make search better?" This assumes that one size fits all and that we can design one correct way to experience information. A look at any magazine rack quickly tells you that content needs vary dramatically and that people like to experienee information in many different ways. Harvard Business Review, People, and Wired each represents a different aesthetic, appropriate to the content and the customers they reach.

So what about search? Should the search results for "Paris Hilton," "hematoma," and "chicken recipes" all look the same? We think not. "How can we make your search for [name a topic] better?" is a fundamentally different question from "How can we make search better?" By starting with a specific search query (and by designing for a specific intent), we're continuously uncovering new ways to make specific search experiences better. Think of this as custom-tailored search results.

We start at the top of the User Experience Hierarchy of Needs pyramid, asking, "What kind of search experience is appropriate for this specific search?" Then we design that screen. This has led to at least 18 unique ways to view search results, with hundreds more planned. What's interesting about this approach is just how different the search results can be, in terms of both the presentation of information and the data sources we choose to aggregate. As you'll see, sometimes searching the entirety of the Web doesn't produce the best results.

Mmm, good