Yale professors Judith A. Chevalier and Dina Mayzlin (Economics and Management, respectively) have released the results of a study they did analyzing the effects of "word of mouth" (in the form of customer reviews) on book sales at Amazon and BN.com. (You can also read a draft of their final paper here.)
Their methodology was not as rigorous as one would ideally wish for (since it's nearly impossible to get hard sales data), but they did come up with some interesting conclusions.
- Overall, reviews are overwhelmingly positive at both sites, and slightly more positive at BN.com.
- Amazon has more reviews, and they are longer and more detailed.
- Prices for the books in their sample were "significantly higher" at BN.com than at Amazon. (This is not apropos to the discussion at hand, but I found it very interesting nonetheless.)
- The addition of favorable reviews at one site increases book sales at that site relative to the other retailer.
- Negative reviews carry more weight with consumers than do positive reviews.
- The impact of a negative review is more powerful in decreasing book sales than a positive review is in increasing sales.
In short, positive reviews boost sales and negative reviews lower sales -- and the effect of the negative reviews is stronger than the effect of the positive reviews.
This was particularly interesting to me since it differs from the traditionally perceived effect of print reviews, where both positive and negative reviews are thought to boost sales. But it just shows the differences in the model when you're looking at a single entity that combines both reviews and sales.
When customers can see the review at the point of purchase, it obviously has a different effect than if they read it in a newspaper, for example, and then subsequently see the book in a store or online. In the latter case, the negative effects of a pan might be more ameliorated, and the book might actually benefit from the increased level of recognition and attention.
If scientists had access to more hard data we could get an even better understanding about the effects of reviews on book sales. (One of the fundamental challenges of publishing is that the publishers don't have the data and economic analysis to support many of their practices.) But this is interesting to see in the meantime.