• by David Berkowitz , April 22, 2008

To search is to discover.

Perhaps there's a Latin phrase for that, or perhaps it is what it is. The intersection of search and discovery came to mind after reading "Jump Point" by Tom Hayes, a thought-provoking snapshot of the near future when there are three billion people connecting online. Hayes references how central discovery has become to the online experience; we'll return to his perspective momentarily.

Discovery online can take a number of forms. It's generally offered as a counterpoint to searching. With search, users enter a query and ideally find exactly what it is they are looking for (even if the search needs to be refined), while discovery presents new information to the user at a time when there's no explicit interest expressed.

Much of the reason for the divergence of discovery and search stems from some people citing how consumers spend only 5% of their time online searching, so the rest of their time online is undervalued (see my rebuttal in the previous column "The Five Percent That Matters"). Yet often searching is the best way to discover something, whether it's a new site, new information, or new products. Yahoo has arguably been the best at making this case in its series of studies on the branding value of search, showing for several verticals and consumer demographics that search rivals word of mouth as the best mechanism for consumers to learn about new brands.

Yahoo plays another important role in the search and discovery overlap with its behavioral targeting offerings, as Yahoo makes it fairly easy to target in-market consumers based on their search activity even when they're not searching. Behavioral targeting in general provides a great framework for how search and discovery work together. It's then a short step from there to Google's efforts with personalized search (refer to the column "Optimizing for One or 12 Trillion" for more). After all, personalized search results are adapted to a searcher's previous behavior to better prioritize which links a user will discover, and it falls under the "target people, not publishers/pages" mantra that behavioral targeting popularized.

Discovery presents a form of organized chaos, which Hayes describes: "Whereas great fortunes have been amassed by the Yahoos and Googles because they brought order to the linear world of Web search, the emphasis is now on discovery, not search. Discovery is a decidedly nonlinear process. When consumers adopt a discovery mindset, contextual, see-and-say advertising strategies break down. The online shopper who follows a link to a jewelry offer may be just a few clicks away from buying a car."

Companies pioneering the discovery mechanisms have a tough balancing act. They need to foster those "aha" and "wow" moments on the Web, while still making the consumer feel like he or she is the one discovering it. From the consumer's perspective, discovery engines can even seem bizarre. This was evident in a review of Aggregate Knowledge, one of the most established companies in the discovery space, that Richard MacManus posted on Read/Write Web. MacManus said, "I clicked some stories on the BusinessWeek.com homepage, and noticed a ‘More from BusinessWeek' list of links to the right of each story. However, none of these links seemed very relevant to the story... Based on my tests, it doesn't seem like there is much - if any - semantic analysis of the page content in order to come up with the ‘More from BusinessWeek' links."

Aggregate Knowledge CEO Paul Martino responded, "Different and serendipitous results happen as a result of this optimization. For example, many results might end up going into a new category of stories that are not from the category of story being looked at... The objective is to continue to allow the user to discover new and fresh content." And then MacManus, coming around but still not sold, commented back, "I think it would help if AK told us a little more about how the algorithm works -- there's a fine line between randomness and serendipity, and right now it's hard to see that line."

Perhaps MacManus can't see the line because it doesn't exist. Consider the consumer's perspective: some of the links will be random, while some will be serendipitously on the mark. With search, there isn't as much serendipity for the most directional searches, but those flashes of excitement come often enough when the results present the consumer with unexpected directions.

The blurring of the line between search and discovery means that the line between direction and serendipity will blur in turn. That should be fine with consumers if they get what they want, whether they expected it or not.