Arro does something similar (I'm a founder). Scan a product's barcode, it analyzes several million reviews and shows better choices: http://arroapp.com
How do relationships with the sources referenced work? Is consent given? I can certainly imagine some valuable sources politely declining to be scraped.
I don't know, I see this as potentially cool, but really very spammy.
Think about how trivial something like this is to manipulate, and if we're talking about the value of consumer electronics going into the hands of our nations youth; well that doesn't inspire me with confidence. If there's potential to gain, there will be potential for exploitation (see Cory Doctorow's excellent piece 'All Complex Ecosystems have Parasites'[0])
I don't think you can monetize uneducated commentary. This would be a lot more interesting, to me, if the conversations were between experts and the points were upvoted by the community. I see value in moderated debates with scoring provided by the community. I don't see value in anyone spouting their opinion and having upvotes on each side.
It's like reddit but with consumer sentiment at stake.
You are absolutely correct. Most review systems which are owned by the merchants are largely viewed as a mechanism for increasing conversion. We don't make money by selling things though.
So although attempts to game the system will abound we are expressly and only incentivized to create a pure community with a genuine dialog. If the users lose we lose. Given incentive that strong we will have to confront gaming head on rather than be complicit.
I had an idea a few years ago. Instead of getting people to come to a new site, why not harvest all the data on existing enthusiast sites out there? They have alllll the info on what model years are good and which to avoid, etc.
Epinions has definitely got it's place but we are going for a community rather than a paid long-form review destination. We feel like human beings talking conversationally about the things they love is more accessible and interesting.
When I read the headline, I imagined a service that crawled forums and discussion boards and used NLP to determine useful shopping advice.
If that happened, all those non-slept nights spent battling people who were wrong on the Internet would finally pay off. (http://xkcd.com/386/)