I like StumbleUpon, the website recommendation engine that lets me click a button whenever I’d like to view a new, potentially interesting website. But for all the hours it has helped me waste, I wouldn’t go as far as to call it useful – pages that get recommended are rarely related to each other beyond a general category, so there’s never any logical train of thought.
San Francisco-based startup BuzzBox is looking to add some logic to the art of ‘stumbling’ with its new Firefox plugin, Fast Forward. The service generates site recommendations based not only on their popularity, but also by the order in which they were viewed. For example, when I clicked the ‘Fast Forward’ button while reading about Twitter users reporting on the terror attacks in India, the service directed me towards the latest CNN coverage on today’s atrocities, as this was the page most frequently visited after reading the TechCrunch article.
To accomplish this feat, Fast Foward records and analyzes the anonymized browsing habits of its users. CTO Mike Prince assures me that all data is totally anonymous and that the service ignores any secure browsing (like banking or Email), but there are still a few possible issues. For one, users may be inadvertently directed to staging pages that are typically obfuscated (I actually managed to land on a BuzzBox alpha page by hitting Fast Forward from its homepage). Prince acknowledges that there are still some issues, explaining that it is still in an experimental stage. And if you do frequent sensitive sites, it’s easy to turn off tracking entirely.
While the plugin is for Firefox only at the moment, the company plans to have an Internet Explorer version available in the next few months. For now BuzzBox isn’t doing anything to monetize Fast Foward, but in the future the plugin may integrate unintrusive overlays at the top of some sites with small ads (Prince emphasizes that the plugin won’t be annoying).
I like Fast Forward – in my testing it was usually easy to tell why certain pages were being recommended, and while it may not always be as entertaining as StumbleUpon (because the recommendations are less varied), I find it much more useful. That said, the service will need to seriously ramp up its privacy settings, otherwise paranoia may keep its userbase small.