• by Cory Treffiletti , Featured Contributor, March 19, 2008

Last week I was introduced to the idea of visual search. If you haven't heard about it yet, then take special note, because I think this will be important.

Visual search is the concept of reorganizing and representing search results in a graphical manner -- rather than text-based results in the way Google re-engineered back in the late '90s.

The example I saw was for a new company called SearchMe. If you go to its site, you can view a short video that provides examples of what its interface looks like. SearchMe owes the entirety of its design to iTunes, making use of the same "coverflow" view that Apple pioneered for displaying album covers and numerous other sites mimicked in a matter of days from its release. The visual search interface returns a series of cover shots that show the first page of the Web site as well as a brief description. The typical text results are found at the bottom of the page, and both elements of the site can be resized to match your viewing tastes.

The returns allow you to see the sites that match your inquiry, providing the user more insight into the sites before they click through to view them. Assuming the search algorithm is accurate and at least as effective as that of Google or Ask.com, then this interface appeals to me -- as there are many cases when the text results for a site may be accurate, but the quality of the overall site design can translate to a sub-par experience. Then I'll revisit the search results and look for another site.

In speaking to a venture capitalist last year, as I was working on my own search interface for a site that I theorized might be appropriate to the next stage of search, I was told that any new innovations in search should allow the user to find what they're looking for through faster means. I'd been focusing on the layout of the results and trying to incorporate Web 3.0 elements of artificial intelligence into the algorithm, thereby hypothesizing the types of information that an individual user might find of interest based on behavioral data and past interactions. That model may just have been too complex -- and I might have simply overlooked the concept of a visual search display. A visual search display does what that VC asked for; it provides faster access to relevant information. It's a simple concept and one that can easily be overlooked.

Of course there is one potentially large challenge for visual search -- which is that it requires more bandwidth than a traditional text-results page. But bandwidth shouldn't be an issue for very long. We see that video is no longer burdened by bandwidth, and more people are accessing richer online content from their homes. At work bandwidth is already pretty fast, and across the board we know that Internet access speeds are only going to get faster over the next two to three years. The only place where visual search may not be appropriate is on a mobile platform, where the traditional text results are more easily accessed from SMS. That being said, mobile platforms are getting stronger -- and the barrier here was also access speeds, but that issue will be revisited over the coming years, as well.

The barriers and challenges to visual search exist, but they're finite -- whereas search isn't. There really hasn't been much innovation in search that caught on to the mass market, because Google is a huge player, more than an 800-pound gorilla. That being said, if it saves time and it's visually appealing, then consumers will try it out -- and trial can lead to viral, and viral can lead to mass penetration.

It remains to be seen whether SearchMe is the next stage of evolution for search but I think it's off to a great start. I think the concept of visual search will be a buzz term for the later part of this year.