I’ve already my outlined my point of view on online reviews — flaws in review-based systems are starting to become evident and we need to think about ways to design to match new reviewing behaviors.

One flaw in reviews is the phenomenon of blind reviewing. Blind reviewing occurs when someone posts a review about something they don’t actually have any experience with. Sometimes this is malicious, sometimes its not.

This is especially common when posting a review is the only way to share information on a site. The other day I was looking a restaurant review on Yelp. 35 people had reviewed the restaurant well, but the most recent reviewer decided to give the restaurant 3/5 stars, since the restaurant was closed for the month when she went.

That is useful information, and it saved me some disappointment, but her review is just one of many written by people that haven’t experienced the things they are reviewing. When I was in Austin recently, I almost went to a restaurant based on positive Yelp average reviews. Turns out most of the reviews were for a nice bartender that reviewers knew socially….most of the reviewers had never actually dined there.

All this may seem minor, but when reviews are on a five point scale, every rating counts in terms of search result placement.

The solution:
Let people submit “Tips” — Let them add valuable information to sites without forcing them to make a review. Introduce new content types (e.g. Tips) that allow users to add valuable information. Display this information differently and create a Wikipedia-esque/Digg-like culture of collaboration. Give users the opportunity to agree or disagree.  This encourages further submission of valuable content, adds clarity to reviews, and gives people something to do once they’ve already reviewed something.

Amazon has started to tackle this problem, but the execution is complex and probably not a good model to work from. Keep it simple and I think good things will come of this direction.