It's an interesting concept. I believe that there is a lot of untapped potential in social networks and discussion boards, information that is not going to be tapped easily by search engines.
Relevant mind is offering an alternative to the product reviews you can already find all over the internet. The problem with product reviews is that there's a lot of "review spam" out there. The phrase was coined by Nitin Jindal and Bing Liu at the Department of Computer Science, University of Illinois at Chicago. Some reviews are written by the people trying to sell the product, or by competitors panning a product, although the researchers were not able to come up with a good method for determining which or how many reviews were false.
But let's face it, the book reviews at Amazon.com, for example, usually start with reviews written by friends of the authors. They're good-hearted but misleading attempts to help friends sell their books. CEO Aaron Mann figures that 1/3 of the reviews on Amazon qualify as spam reviews.
Even sites that let you rate a product, such as Netflix, have trouble getting real results. Every product tends to skew toward average, a rating of 3.7. Sites like Netflix have to tweak their rating algorithms with such techniques as giving more weight to negative reviews.
If you're a real aficionado of a product you might join a discussion or hobbyist group. But how to find such groups when you're first starting out?
Relevant Mind tries to help that problem. You can search for particular products, brands, or just click on a price range. When you find a product you might be interested in, click on it and a tab of "Conversations" will appear. Those conversations are brief excerpts from discussions of the particular product on the internet. Click on an excerpt that looks interesting, and you are sent to the site where the discussion appears.
This takes a lot of work, both in coming up with a comprehensive list of products and in finding the relevant discussions. So Relevant Mind is starting with two categories, road bikes and golf equipment, since those are strong interests of two of the founders.
Other categories -- sports, cooking and musical instruments are on the to-do list -- will be added by hobbyists who want to make money providing information about their favorite pastime. It takes expertise to do this right. Relevant Mind will be seeking out and hiring such people through blogs and user groups. "I'm most looking forward to hiring the Vertical Gurus," says CEO Mann. "They'll be the funnest people at our company."
The company will avoid product categories where there are already good independent reviewers, such as consumer electronics. You can go to CNET or any number of newspapers for that.
The site accepts no ads, but tries to make money by linking to retailers who have affiliate programs paying fees for sending people their way. Click on the "Buy now" tab and you'll get a link to sites selling both new and used products.
You'll also be able to create a personalized page on the site or in your browser to help the company learn your likes and dislikes and give you more relevant results.
Here's the difficulty. Some retailers may have great prices but no affiliate program. CEO Mann says he will send people to those sites anyway in the interest of making the site truly useful to buyers. You can't buy new road bikes online, for example, so the site tends to link to people selling them used. That's the right idea, but it will take a lot of willpower to stick to the ideal.
It's an incredibly labor-intensive business model, which is one of the advantages. As long as Relevant Mind manages to develop comprehensive product categories and link to good discussions, it will have a huge first-to-market advantage. This is not something Google will be able to do.
It looks interesting. The site has some bugs to work out. When searching for the cheapest bikes, the results tend to come up with a lot of components rather than complete bikes. But if the site manages to come up with truly relevant results, including finding good retailers, it has promise.
There's a lot of intelligence residing on the web. It's just a matter of helping people find it.