I tried out Grokker today, the latest release from Yahoo! offering a "visual search" tool that groups search results contextually into circular clusters. Despite a somewhat clunky visual interface with some strange mouseover behaviours, it definitely has promise for giving users a more intuitive way to use online search more powerfully to hone in on the most relevant results for what a user is seeking. Another promising visual search tool, Newsmap has what I find to be a more intuitive view focused more on the content and less on graphics. A quick list of some key trends in enhancing search I’ve noted just over the few months includes:
- Advanced Sorting – at first only the realm of online retail sites which offered users the ability to sort search results based on price, product, brand, discount, etc … now this tactic is extending across many retail and non-retail sites such as online travel aggregator sites (Kayak and Sidestep), and real estate sites. Yahoo has a beta release of a service termed Mindset which offers the ability to sort by intent, from purchasing to researching. Clearly a powerful way of helping the user get the most relevant results, and setting them up nicely to charge a premium of keyword advertising through Overture for those sites that appear highly on the purchasing side of the scale. Smart.
- Visual Grouping – taking search results and putting them into clusters to help users hone in on the type of information they are seeking. Examples of sites like this include Yahoo’s Grokker, The Hive Group’s Newsmap and Tagcloud, a folksonomy based blog search tool that allows you to cluster by keyword mentions and get an understanding of relative buzz words within a search context at the same time. (Now Technorati has something similar too) My only problem with the Grokker idea is that there doesn’t seem to be a relative scale to illustrate importance of clusters. For typical searches for a single source, this may not be an issue. But for research-style searches where you are trying to get a picture of the landscape online around a particular issue, visual grouping search has so much more promise. Perhaps this may be something they are considering adding to future releases.
- Self-Learning Search – I first experienced these types of searches a number of years ago when we were implementing a notes-based intranet at Didata. It was the most promising search technology at the time — learning search, that could refine itself based on what you clicked on or noted as relevant and not relevant. Unfortunately, it usually forces an extra step from users in order to work, the kiss of death for innovation in an ease-of-use driven space.
- Niche files/mediums – I love Google images. It’s saved me lots of research time over the past year in finding images that I need quickly, from client logos for proposals to resource imagery for presentations. With Google’s new video search tool, the range is complete. I can now go online and use Google to hone in on files in PDF, PPT, DOC, images, and video. Maybe in the future I can go and search for PSDs. But then I guess that would make Google a facilitator of copyright infringement, wouldn’t it?
So what potential do all of these innovation trends offer for search engine marketers? My guess is that visual search and other such innovations will offer marketers a greater ability to reach users not only based on their activity, but on their intent. And yes, we will be paying a premium for it.
1 thought on “Visual Search and the Future of Search Engine Marketing”